News & Updates
The latest news and updates from companies in the WLTH portfolio.
How to Spot a Stock Market Bubble 101: Raymond James Just Placed an $800 Price Target on SpaceX, Valuing Elon Musk's Company at $10.5 Trillion
One month ago, on June 12, Elon Musk's artificial intelligence (AI) and space economy conglomerate, Space Exploration Technologies (SpaceX) (SPCX 4.75%), rewrote history with its initial public offering (IPO). The $85.7 billion raised, including the underwriters' overallotment, nearly tripled the previous IPO record holder, Saudi Aramco. But in kicking off IPO mania -- large language model developers Anthropic and OpenAI are expected to follow in SpaceX's footsteps -- SpaceX may also be fueling the final stages of an AI bubble that history suggests is waiting to pop. Rarely are stock market bubble warning signs as glaring as Raymond James Financial's price target assigned to SpaceX. Wall Street's high-water price target foresees SpaceX reaching $800 in 2031 Given that 21 underwriters helped bring SpaceX public and received shares for doing so, it should come as no surprise that Wall Street analysts have, as a whole, presented an overwhelmingly positive outlook for the company. But Raymond James Financial analyst Brian Gesuale is a true outlier. His $800 price target by 2031 implies 451% upside, based on where SpaceX's shares ended on July 10, and assumes a valuation of roughly $10.5 trillion. For context, this would be more than double Nvidia's current market cap. Gesuale foresees SpaceX's full-year sales scaling from an estimated $38.5 billion in 2026 to approximately $837 billion by 2031. More importantly, earnings before interest, taxes, depreciation, and amortization (EBITDA) are projected to catapult from $17.7 billion in 2026 to $696 billion by 2031. While there's no question that AI and the space economy are two of the hottest addressable opportunities on Wall Street, several headwinds suggest Gesuale's pie-in-the-sky price target is pure fiction and the sign of an end-stage bubble that's about to burst. SpaceX spotlights everything wrong with Wall Street Although the stock market is a long-term wealth-creating machine, it's prone to occasional bubble-bursting events. SpaceX's current $1.91 trillion valuation and Raymond James' $800 price target for the company spotlight everything that's wrong with Wall Street over the short term. For starters, SpaceX hasn't demonstrated that its operating model is sustainable. While satellite-based broadband services provider Starlink is profitable, AI start-up xAI -- the segment responsible for the lion's share of SpaceX's $28.5 trillion addressable market -- is burning cash as Musk's company chases AI compute capacity. Elon Musk also has a terrible track record of fulfilling lofty promises and innovative expectations. As CEO of Tesla, Musk proclaimed that 1 million robotaxis would be on public roads by the end of 2020, which never happened. He's also assured investors that Level 5 full self-driving is "one year away" annually for more than a decade. Musk continually overpromises and underdelivers. SpaceX is likely to be haunted by historical precedent, as well. No company at the forefront of a game-changing technology has sustained a price-to-sales (P/S) ratio above 30 for any extended period. SpaceX is trading at roughly 50 times Gesuale's forecast sales for this year. Lastly, every game-changing technology for more than three decades has navigated an early stage bubble-bursting event. These bubbles have formed because investors constantly overestimate the optimization timeline of innovations. It'll likely be years before SpaceX's solutions are optimized, making Raymond James' high-water price target highly unlikely.
LTM Partners with Anthropic to Expand Enterprise AI Adoption
LTM has partnered with Anthropic to integrate Claude into its AI platform, helping enterprises scale AI adoption across software development and business operations. LTM has partnered with AI company Anthropic to help enterprises deploy Claude across software engineering, application modernisation, and business operations, as organisations increasingly shift their focus from AI experimentation to large-scale implementation. The partnership will see Claude and Claude Code integrated into LTM's BlueVerse AI Delivery Fabric, the company's enterprise AI platform, to support AI-assisted software development, application modernisation, site reliability engineering (SRE), observability, and agent orchestration. The collaboration is expected to initially focus on industries including banking and financial services (BFSI), technology, consumer businesses, and manufacturing. Helping Enterprises Scale AI Deployments While many enterprises have experimented with generative AI over the past two years, scaling those initiatives across business functions remains a challenge. LTM said the partnership aims to address this gap by combining Anthropic's AI models with its enterprise implementation capabilities. Chris Ciauri, Managing Director of International at Anthropic, said the partnership is intended to help customers embed AI into their software development and modernisation programmes. "LTM brings delivery expertise, trained professionals, and long-standing client relationships across industries. By embedding Claude and Claude Code into BlueVerse, the partnership enables customers to integrate frontier AI into how they build, modernise, and manage software," he said. Building Skills Alongside Technology As part of the collaboration, LTM will expand its AI1000 talent initiative to train architects and engineers on Claude, while also establishing a dedicated Center of Excellence (CoE). According to the company, the CoE will focus on developing reusable AI skills, reference architectures, governance frameworks, and best practices for deploying AI across enterprise environments. It will also support areas such as responsible AI, model governance, and data privacy. Enterprise AI Moves Beyond Pilots LTM said it will also integrate Claude into its own delivery operations to establish internal AI workflows before extending those capabilities to customers. Venu Lambu, CEO and Managing Director of LTM, said the partnership is aimed at helping enterprises translate AI investments into measurable business outcomes. "Combining Claude with our BlueVerse platform, domain expertise, technology capabilities, and AI1000 talent initiative creates a foundation for enterprises to embed AI across their businesses and accelerate modernisation," he said. As enterprise AI adoption matures, technology service providers are increasingly partnering with model developers to deliver implementation, governance, and industry-specific solutions alongside AI models. LTM's collaboration with Anthropic reflects this shift, with the emphasis moving from access to AI models to deploying them at scale across business operations.
Pete Hegseth Called Anthropic a 'National Security Risk.' Now CISA Is Using It.
CISA is one of around 100 organizations approved to use Anthropic's Mythos model, which is far better than human penetration testers at detecting and exploiting software bugs. The United States Cybersecurity and Infrastructure Security Agency (CISA) is now using artificial intelligence (AI) developer Anthropic's Mythos platform to "audit government software" -- a seeming rebuke of the Pentagon, which classified the company as a "national security risk" in February and attempted to institute controls on its use by government agencies. CISA's Attack Surface Evaluation team has been employing the model to audit source code and identify flaws that could be exploited by cybercriminals or nation-state actors, Reuters first reported last week. The initiative is apparently part of a pilot program to evaluate whether AI can accelerate software security reviews across government systems. According to the Reuters report, Mythos successfully identified multiple vulnerabilities during testing, although specifics on the number of vulnerabilities, severity, or affected software were not disclosed. What to Know About Claude Mythos The Claude Mythos platform is an extremely advanced large language model (LLM) that hasn't been released to the general public due to concerns that its software engineering capabilities could be too dangerous. The software has already been used to identify thousands of zero-day vulnerabilities in every major operating system and web browser, including decades-old bugs. As the AI can spot even the most subtle bugs, it has the capability to self-correct any coding mistakes, giving it clear commercial applications. The main risk is that in the wrong hands, it could also autonomously execute exploits at scale and speed that far exceed any human capabilities. "Software code review and analysis are nothing new. Realistically, most issues found are not exploitable without very specific conditions being met (i.e., the vulnerable function needs to actually be invoked and exposed to the attacker in order to be abused)," explained Chris Traynor, penetration tester at Black Hills Information Security (BHIS) and instructor at Antisyphon. Traynor told The National Interest via email that he believes that AI vulnerability scanning will likely find many new and novel issues that were simply too complex to identify with legacy tools before. "But added complexity can cause limitations in exploitability," Traynor said. "AI scanning will likely produce a lot of unactionable output very quickly that will need to be reviewed by experts to find the real risks." AI Is Far Better Than Humans at Finding Bugs -- for Better and Worse Software bugs are virtually inevitable and incredibly common in development. Complex systems require thousands of lines of code, and anything from minor typos to miscommunications to logical missteps can lead to bugs. This introduces unpredictable variables. Code also doesn't exist in a bubble, nor is it created in one. Changes in operating systems, web browsers, and third-party services can break functioning software, often without warning. More worrisome is that such software bugs serve as the foundation for almost all exploits in cybersecurity. A bug has the potential to create unintended weaknesses, such as how a program handles data or permissions. Many software vulnerabilities can go unnoticed, often hidden for years. The danger is that AI can find those quickly. It is an extremely useful, but essentially neutral, tool: a developer or "white hat" penetration tester can use AI to help patch faulty code, or a "black hat" hacker can use it for nefarious purposes. "AI finding vulnerabilities in federal code at scale is interesting, but the harder question is what happens after the finding. A vulnerability that exists in a library no one calls, behind a network segment no one reaches, is not the same problem as one sitting in a critical authentication path," said Seemant Sehgal, founder and CEO of cybersecurity provider BreachLock. "Without validating exploitability and reachability, every finding lands with the same weight, and that creates its own kind of risk," Sehgal told The National Interest. "The real test of this program is whether the output helps prioritize action or just expands the backlog." The US Government Has a Mercurial Approach to AI The federal government is already using AI extensively, deploying more than 3,600 distinct AI programs across its agencies. But the Trump administration is less clear about how it believes AI should be used in security matters. "The federal government can't seem to decide what it thinks about AI in general, or Mythos in particular," Bronwen Aker, AI research and strategy analyst at BHIS, told The National Interest. "One week Anthropic is a supply-chain risk, the next week CISA is handing Mythos the keys to scan federal code for vulnerabilities." The intra-government squabbles over Anthropic are by now a matter of public record. In February, the Pentagon called for Anthropic to be banned across the entirety of the federal government. Four months later, in June, the Department of Commerce instituted an export ban on two powerful Anthropic models, Fable and Mythos, leading Anthropic to shut both of them down. Earlier this month, the Department of Commerce lifted the export ban, and Fable has been made available to the general public once more. Mythos had been strictly limited to just around 100 vetted US organizations focused on critical infrastructure and cybersecurity. That seems to include CISA, which is employing the platform in the software audit. "That inconsistency would be bad enough to start with, but because it's not clear what Mythos is actually scanning, it's much, much worse," Aker told The National Interest in an email. It remains unclear if it is government-written code or software built by third-party contractors and vendors. "In-house bugs are one problem," warned Aker. "Vendor bugs running across federal systems are a supply chain problem, and the public has a right to know which one this is." AI-Generated Code Is Part of the Security Problem Part of the problem is that at the same time AI scans for vulnerabilities in legacy code, it also generates new code on the other end -- code with many of the same flaws in it as human-generated code. In other words, using AI to patch bugs is only solving half the problem. Jacob Krell, senior director for secure AI solutions and cybersecurity at security provider Suzu Labs, observed that CISA pointing Mythos at government codebases should be seen as a smart move, but perhaps only the opening move in what could be a long game. "I've seen federal systems running code that hasn't had a serious security review in a decade, and a model like Mythos can cover that volume in hours instead of months," Krell told The National Interest. "Every federal agency and contractor also has developers writing code with AI assistants, and those tools produce insecure output more often than secure output. Authorization flaws, hardcoded credentials, missing input validation, all shipping by default because the models optimize for 'does it run' and skip 'is it safe.'" The danger is that when combined, the result is akin to a treadmill that isn't so easy to get off of. In this case, Mythos finds legacy bugs, teams patch them, but then AI coding tools introduce fresh vulnerabilities even as they fix the old ones. That puts everything from power grids to water systems, which may be privately run, still squarely in the crosshairs of hackers. Krell suggested that CISA can't simply harden federal code and call it done. "If the agency has a scanning tool this capable, the operators running critical infrastructure need access to it too, because those are the systems that actually keep the lights on," he said. "I'd want CISA to pair this initiative with secure-generation standards for AI coding tools in federal development, and extend scanning access to critical infrastructure operators. We are draining the pool while the hose is still running." About the Author: Peter Suciu Peter Suciu has contributed to dozens of newspapers, magazines and websites over a 30-year career in journalism. He regularly writes about military hardware, firearms history, cybersecurity, politics, and international affairs. Peter is also a contributing writer for Forbes and Clearance Jobs. He is based in Michigan. You can follow him on Twitter:@PeterSuciu. You can email the author: [email protected].
Samsung will make 2nm AI chips for Anthropic
A few days ago, it was reported that Samsung could win a contract to manufacture high-performance AI chips for Anthropic. A new report now claims that Samsung Foundry has secured the contract to make chips for Anthropic, the AI firm behind the popular Claude AI system. According to Newsworks Korea, Samsung Electronics has secured a major contract to fabricate proprietary AI accelerator chips for Anthropic. The chip will be fabricated using Samsung Foundry's cutting-edge 2 nm process node and packaged with the company's advanced packaging technology. It is unclear whether the chip will be manufactured at Samsung's plant in South Korea or its facility in Taylor, Texas. It is worth noting that Samsung Electronics invested in Anthropic's $65 billion Series H funding round. Anthropic is a US-based AI firm founded in 2021 by ex-OpenAI researchers, and it competes directly with OpenAI. Anthropic is significantly more valuable than OpenAI. Samsung Foundry's reported Anthropic contract comes just a few months after it secured chip orders from Tesla and Groq. This could help the company return to profitability next quarter. Samsung is also reportedly in talks with Google and Meta to manufacture AI chips, or components for their AI chips. Due to the ongoing AI boom, TSMC, a much larger contract chip manufacturer than Samsung, is facing significant production backlogs. As a result, it is charging high prices to make chips and is reportedly prioritizing orders from major brands such as Apple, AMD, MediaTek, and Nvidia. This could allow Samsung Foundry to attract more customers. Samsung Electronics is currently the only company that offers chip fabrication, chip packaging, and memory chips under one roof.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta on Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTX, and XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens and $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per month, or $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Conclusion Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta on Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5et Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTXet XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens and $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per month, or $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Conclusion Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta on Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5a Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTXa XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens a $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per month, or $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Závěr Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta on Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5, y Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTX, y XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens y $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per montho $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Conclusión Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta on Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTX, and XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens and $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per month, or $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Conclusion Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta auf Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5und Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTXund XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens und $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per month, or $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Schlussfolgerung Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta on Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTX, and XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens and $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per month, or $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Conclusion Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Claude Design: Anthropic's AI Tool (Access, Pricing, Uses)
Claude Design is Anthropic's prompt-based design workspace, and more than one million people used it in its first week. It turns plain English into editable slides, interactive prototypes, mockups, and one-page assets, then exports them to PDF, PPTX, Canva, standalone HTML, or straight into Claude Code. It launched on April 17, 2026, lives at claude.ai/design, and on launch day Figma's stock fell about 7%, according to Gizmodo. Claude Design got a major overhaul on June 17, 2026 that changed how it works and what it costs you in usage. It now imports your real design system, round-trips with Claude Code through a /design-sync command, lets you drag and resize elements directly on the canvas, and draws from your shared Claude usage pool instead of a small separate allowance. You also now pick which Claude model does the work. This guide covers what Claude Design is, how to access it on Mac, Windows, and Linux, what each plan costs, what you can build with it, and how it stacks up against Figma and Canva. The Key Takeaways * Claude Design is in beta on Pro, Max, Team, and Enterprise plans and included with your subscription. There is no free tier, and Enterprise has it off by default. * The June 17, 2026 update added design system imports, a two-way /design-sync round trip with Claude Code, direct canvas editing, and brand lockdown controls for admins. * Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork. The separate design allowance that users burned through in minutes is gone. * It is browser only on Mac, Windows, and Linux. There is no native Claude Design app, and any third-party ".dmg" or ".exe" download claiming to be one is not from Anthropic. * You now pick the model inside Claude Design, choosing between Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5, alongside an Effort setting. The "powered by Opus 4.7" line from the April launch post is out of date. What is Claude Design? Claude Design is a prompt-based design surface built by Anthropic Labs, the team inside Anthropic that ships experimental products. You describe what you want in plain English, and Claude generates a working design on a side-by-side canvas. You refine it through chat, inline comments, direct manipulation, and layout controls that adjust spacing, color, typography, and alignment. The product is documented in Anthropic's launch announcement. It launched on April 17, 2026, three days after Anthropic CPO Mike Krieger stepped off Figma's board, a resignation TechCrunch dated to April 14. It arrived as a research preview and moved to beta with the June update. At launch it ran on one fixed model, Claude Opus 4.7, which Anthropic then called its most capable vision model. That is no longer how it works. Claude Design now ships a model picker, and you choose the model yourself. Claude Design accepts a wide range of inputs. You can drop in screenshots, photos, DOCX, PPTX, and XLSX files, or point it at a GitHub repository so Claude reads your real components. There is also a web capture tool that grabs elements directly from your own website, so prototypes look like the real product. Anything you build can be shared inside your organization with view-only, comment, or edit permissions. Who is Claude Design for? The tool serves three core audiences. Founders and product managers who need a clickable prototype before they brief a designer. Sales and marketing teams who need a polished pitch deck today, not next week. Engineers who want to sketch an internal admin UI without leaving Claude. Designers themselves are not the primary target, and Claude Design complements tools like Figma and Canva rather than replacing them. What changed in the June 2026 Claude Design update Exactly two months after launch, Anthropic shipped a major Claude Design update on June 17, 2026. The theme was brand consistency and getting designs into production code, and it also fixed the tool's most complained-about flaw. The full changelog is in Anthropic's June 17 announcement. Here is what actually matters. Design system imports Claude Design can now pull in one or several design systems from a GitHub repo, design files, or raw uploads. Once a system is in, Claude builds with your real components, checks its own output against your system, and corrects itself before you ever see the result. This is the difference between a mockup that looks vaguely like your product and one that uses your actual buttons, cards, and tokens. The /design-sync round trip with Claude Code The headline feature is two-way traffic with Claude Code. Running /design-sync inside Claude Code pulls your local codebase's design system into Claude Design, so every prototype starts from existing components instead of approximations. A /design command in the Claude Code terminal lets developers create, edit, and sync design projects without leaving their workflow. The trip back is just as short. When a design is ready, it hands off to Claude Code, which picks up where you left off with no screenshot and no rebuild. If you are already living in the terminal, our roundup of the best Claude Code skills covers what else you can bolt onto that workflow. Direct canvas editing Claude Design used to route almost every tweak through the chat pane. The update added rich layout controls that let you drag, resize, and align elements directly on the canvas. You still ask Claude for the big structural changes, but nudging a card two pixels left no longer costs you a prompt. Brand controls for Team and Enterprise Larger organizations get a new admin role that can approve a single standard design system and lock down edits. Every asset Claude produces then conforms to company guidelines, which is what turns Claude Design from a prototyping toy into something a brand team will actually sign off on. Shared usage limits, the token-burn fix This is the change that matters most to your wallet. At launch, Claude Design metered usage from its own small allowance, and it drained fast. PCWorld reviewer Ben Patterson burned 80% of his weekly Claude Design allowance in roughly 25 minutes on the Pro plan, producing three variations of a single webpage prototype. His review was headlined "I tried Claude Design for half an hour. I'm already locked out for a week." Anthropic scrapped that model. Claude Design now shares usage limits with chat, Claude Code, and Claude Cowork, so there is no separate design allowance to track and you draw from one larger pool. Anthropic also says it cut average token consumption per turn while holding output quality, and that error rates dropped sharply. Which model powers Claude Design? This is the question almost every other guide still gets wrong. Anthropic's April launch post said Claude Design ran on Claude Opus 4.7, and that page has never been updated, so write-ups keep repeating it. Open the product today and it is plainly false. Claude Design carries the same model picker as the rest of Claude, and Anthropic's support documentation lists Claude Design among the surfaces where organization model-access settings apply, right next to chat, Cowork, and Claude Code. The model picker: Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5 Open the Model dropdown in the prompt composer and you get four models. Fable 5 is pitched for your toughest challenges, Opus 4.8 for complex tasks, Sonnet 5 as the efficient everyday pick, and Haiku 4.5 for fast answers. A More models row opens the rest of the catalogue. No single model "powers" Claude Design any more, so the honest answer is that you decide, per project. In practice, Sonnet 5 is enough for most slide decks, one-pagers, and simple mockups, and it is the cheapest way to iterate. Step up to Opus 4.8 or Fable 5 when you are building a dense interactive prototype or working against an imported design system, where the model has to hold a lot of structure in its head at once. Drop to Haiku 4.5 for quick throwaway variations. The Effort control Underneath the model list sits Effort, which defaults to High. Effort trades intelligence against speed and usage, and Anthropic documents the full ladder as Low, Medium, High, Extra high, and Max in its guide to model, effort, and thinking settings. Because design work now draws from your shared pool, turning Effort down on rough drafts is the single easiest way to stretch a subscription further. Admins get a say too. Since July 1, 2026, Enterprise administrators can restrict both which models their users can reach and which effort levels they are allowed to set, which is the same rollout that brought the picker into Claude Design in the first place. Fable 5 and the included-usage window Claude Fable 5 launched on June 9, 2026 and sits above Opus 4.8 as Anthropic's most capable widely released model, with a 1 million token context window and API pricing of $10 per million input tokens and $50 per million output tokens. That capability is not free-flowing. Anthropic includes Fable 5 on paid plans for up to 50% of your weekly usage limit, after which you switch models or buy usage credits, as set out in its Fable 5 redeployment note. That inclusion window is promotional and Anthropic keeps extending it, first to July 7, then July 12, then July 19, 2026. The picker prints the current deadline next to Fable 5, so check the label before you commit a long session to it. Free plans never had access, and standard Enterprise seats are excluded. How to access Claude Design Claude Design is a browser-based product at claude.ai/design. It works identically on Mac, Windows, and Linux because all generation happens on Anthropic's cloud. There is no native Claude Design app for any operating system and no separate desktop installer. Here are the four steps to get in. Optionally, run /design-sync in Claude Code first to import your team's design system, so generations stay on-brand from the very first prompt. Once you are in, you write a prompt in the left chat pane, watch the design build on the right canvas, then fine-tune with the layout controls or by dragging elements directly. Using Claude Design on Mac, Windows, and Linux Because the heavy generation runs on Anthropic's cloud, any Apple silicon Mac, from the original M1 right through the current M5, runs Claude Design identically to a Windows laptop or a Linux desktop. You do not need a discrete GPU, and the Mac mini, MacBook Air, and entry-level MacBook Pro all handle the canvas without lag. The same goes for an average Windows ultrabook. If you build dense interactive prototypes, a larger display helps, because the chat pane and canvas sit side by side. Many Mac users prefer to run Claude Design inside the official Claude Desktop app for macOS, which you can download from claude.com/download. It gives Claude a dedicated window away from your browser tabs. The desktop app is not required, but it is the cleanest setup if you work in Claude all day. What you can build with Claude Design Claude Design starts you from five templates, Prototype, Slides, Document, Wireframe, and Animation, or from a blank prompt if you would rather just describe what you want. Each one is refined through conversation afterwards. Below are five concrete use cases with the kind of prompt that gets the best result on the first try. Two of those templates are easy to overlook. Wireframe deliberately strips the output back to low-fidelity structure, which is what you want when the argument is about layout rather than polish. Animation produces motion, and because Claude Design renders slides as HTML rather than PowerPoint, Anthropic's own presentation tutorial notes that it can animate transitions to carry a story across a deck. 1. Pitch decks and quarterly business reviews You describe the deck including target audience, slide count, and key data points, and Claude builds the full deck in PPTX format. Example prompt: "Build a 12-slide Series A pitch deck for a B2B SaaS analytics startup. Include problem, solution, market size, traction, business model, team, and ask. Use a minimal style with our blue and grey brand colors." The output drops straight into Microsoft PowerPoint, Apple Keynote, or Google Slides. For deeper presentation workflows, see our guide on how to use AI in PowerPoint. 2. Interactive prototypes for new features Product teams use Claude Design to make a feature tangible before they brief engineering. Example prompt: "Design a 5-screen onboarding flow for a fitness tracking app. Include a welcome screen, goal selection, profile setup, permission requests, and a celebration screen. Use a light theme with sans-serif type." Because the canvas renders real HTML, CSS, and JavaScript, every prototype is clickable and scrollable in the browser. You share the link with stakeholders, collect comments inline, then iterate without leaving the chat. 3. Landing page mockups and one-pagers Founders without a designer use Claude Design to draft a marketing site or sales one-pager in minutes. Example prompt: "Create a SaaS landing page for an AI invoice tool aimed at freelancers. Include a hero, three feature blocks, a pricing table, FAQ, and a CTA footer. Match the layout conventions of modern fintech landing pages." Uploading screenshots of pages you admire pulls real reference material into the prompt, which keeps the output anchored to actual conventions instead of generic templates. If you already have a site, the web capture tool grabs elements from it so the mockup matches your live product. 4. Mockups built on your real design system This use case got dramatically stronger in June. Import your design system from a GitHub repo, design files, or raw uploads, and a new dashboard mockup will use your actual buttons, cards, modals, and tokens. Claude validates its output against the system and fixes deviations before showing you anything, so designs become drop-in references for the next sprint rather than throwaways. When you are ready to ship, click Handoff to Claude Code and send the design to a local Claude Code agent or to Claude Code on the web. Claude Code picks up the design without starting over, so nothing gets rebuilt from a screenshot. 5. Data dashboards, infographics, and internal tools Claude Design also handles charts, dashboards, and quick internal tools. For text-heavy infographics that demand pixel-perfect typography, ChatGPT Images 2.0 is currently a stronger pick because it specializes in dense text rendering. For interactive dashboards with editable data, Claude Design is the better fit, because the output is real code rather than a flat image. Claude Design pricing and usage limits Claude Design is not free. It is included at no extra cost with Pro, Max, Team, and Enterprise Claude plans. Since June 17, 2026 it draws from the same shared usage pool as chat, Claude Code, and Claude Cowork, so there is no separate design allowance to monitor. Here is what each plan costs, and our full Claude pricing breakdown goes deeper on API and Bedrock rates. Pro: $20 per month Claude Pro is the cheapest way into Claude Design at $20 per month, or $17 per month billed annually. It is the right plan for individual founders, marketers, and engineers who want to try Claude Design without a team contract. Design work now competes with your chat and Claude Code usage for the same pool, so a heavy prototyping session will eat into what you have left for everything else. Max 5x: $100 per month Claude Max 5x costs $100 per month and gives you 5x the Pro usage. This is the sweet spot for power users running daily Claude Design sessions alongside heavy Claude Code or Claude Cowork work, precisely because all three now share one pool. Max 20x: $200 per month Claude Max 20x costs $200 per month for 20x the Pro usage, plus priority access to Anthropic's newest features and models. This is for designers and agencies treating Claude Design as their primary visual workflow. Team and Enterprise Team Standard is $25 per seat per month ($20 annually) and Team Premium is $125 per seat per month ($100 annually). Those Team plans run from 5 to 150 seats. Enterprise is custom-priced through sales and ships with Claude Design turned off by default, so an admin must enable it in Organization settings. Once you exhaust your plan limits, Pro and Max users can switch on extra usage, sometimes called usage credits, and keep working. That overflow is billed at standard API rates rather than being included, so it is a metered top-up, not a bigger bucket. If you bought your subscription through the mobile app, you have to enable extra usage on the web. Claude Design vs Figma vs Canva vs Fello AI These four tools solve overlapping problems from different starting points. Claude Design generates from a prompt. Figma builds on a vector canvas with components. Canva assembles from templates and stock assets. Fello AI sits one level up, giving you Claude, ChatGPT, Gemini, Grok, and DeepSeek in a single app. Here is how they compare on the metrics that matter. The short verdict. Use Claude Design when you need to go from idea to interactive output fast and you already pay for Claude Pro or higher. Pick Figma for production UI work, design systems, and developer handoff at scale. Use Canva for brand-consistent marketing collateral with stock assets. Choose Fello AI at $9.99 a month when you want several frontier models in one app and you do not need a design canvas at all. Fello AI: a multi-model alternative for $9.99 a month If you do not have a paid Claude plan and you mostly need AI-generated slides, documents, and everyday creation work, Fello AI is a cheaper way in. For $9.99 a month it bundles Claude, ChatGPT, Gemini, Grok, and DeepSeek, and its PowerPoint Skill produces real .pptx files straight from a chat conversation. The app runs natively on Mac, iPhone, and iPad with cross-device sync, and it holds a 4.7-star rating across more than 27,000 reviews. The trade-off is the canvas. Fello AI does not give you a side-by-side visual editor, so it is not a replacement for a real design surface. It is the better choice when you want speed and one-app access to several models, and the weaker choice when you specifically need to push pixels. Plenty of people run both, Claude Design for work that needs a canvas and Fello AI for everything else. How Claude Design fits with Claude Cowork and Claude Code Anthropic now ships three Claude products that talk to each other, namely Claude Design, Claude Code, and Claude Cowork. The intended workflow is straightforward. Claude Design produces the visual artifact, Claude Code implements it, and Claude Cowork runs the long-horizon agent tasks on top, like research, document drafts, and analytics. The June update tightened that loop considerably. All three now share one usage pool, one set of credentials, and one billing relationship, and /design-sync plus Handoff to Claude Code are the explicit bridges between visual work and shipped code. Claude Design also sends work out to Adobe, Base44, Canva, Gamma, Lovable, Miro, Replit, Vercel, and Wix, on top of PDF and PowerPoint export. If you want quick presentation generation outside the design canvas, our guide to converting PDF to PowerPoint with AI covers that workflow. Limits and caveats Claude Design is still in beta, and the constraints below matter before you commit a workflow to it. Enterprise organizations have it disabled by default and need an admin to opt in. Anthropic has never corrected its April launch post, which still claims the tool runs on Opus 4.7, so plenty of secondhand write-ups repeat a model claim the live product flatly contradicts. Whichever model you pick, its vision and reasoning limits apply to your designs as well. And because design work now draws from your shared pool, a long prototyping session directly reduces what you have left for chat and Claude Code. On capability, the honest boundaries hold. For brand-heavy marketing collateral with consistent assets across dozens of pieces, Canva still wins. For complex multi-screen design systems with tight pixel precision, Figma still wins. Claude Design is at its best on the first 80% of a design, and most teams will hand the last 20% to a real designer. Anthropic has also shipped Claude Dreaming, its agent memory feature, which is worth watching if you plan to keep design sessions context-aware across weeks. Conclusion Claude Design spent its first two months as an impressive demo with a punishing usage cap. The June 17 update fixed the cap, taught it your design system, and wired it directly into Claude Code, which is what turns it into something you can actually put in a workflow. If you already pay for Claude Pro, Max, Team, or Enterprise, open claude.ai/design and run one of the example prompts above, and run /design-sync first if you have a codebase worth matching. If you do not have a Claude subscription and you want several frontier models in one native Mac app instead, download Fello AI on the App Store. To compare the underlying models powering all of this, see our best AI models guide. If you are weighing the two big AI visual tools against each other, our head-to-head on Claude Design vs ChatGPT Images 2.0 breaks down which one wins for slides and mockups, and our walkthrough on how to use ChatGPT Images 2.0 to create visuals covers the OpenAI side in practice.
Anthropic Says Claude's Values Are Different Depending on Which Language You're Using
Claude Values: Anthropic, founded in 2021 by former OpenAI researchers, has made quite an impact with its assistant, Claude. It first launched in March 2023 and has since gone through several updates, including Claude 3 and Claude 3.5 Sonnet. A recent discovery shows that Claude's values and behaviors change depending on the language used. This inconsistency raises important questions about how we ethically use and function with AI models in multilingual contexts. In March 2024, Anthropic introduced Claude 3 Opus, featuring this extensive context window. This allows the AI to process a far greater amount of information than many rivals. Not only does this enhance Claude's performance, but it also magnifies the implications of its language-dependent values. Claude: Key Details Anthropic's findings about Claude's values being influenced by language come from its internal research. While we don't have all the specifics about the study, the implications are significant. The notion that a language model's ethical framework can change based on the language it uses presents a challenge for both developers and users. For more detail, see OpenAI Blog. Research from places like Stanford University and MIT backs this up, highlighting cross-lingual value inconsistencies as a major hurdle in AI alignment. This suggests that AI models can reflect different ethical perspectives depending on the language they process, making deployment in diverse linguistic settings more complicated. As part of a significant funding partnership, Anthropic secured an incredible $4 billion from Google in 2023. This investment shows the tech giant's commitment to advancing AI technologies, particularly in refining the ethical aspects of models like Claude. The funding aims to boost the research needed to tackle these complex issues effectively. Context The foundation of Anthropic is built on the experiences and insights of its founders, especially Dario and Daniela Amodei, who aimed to create AI systems that focus on safety and alignment with human values. The challenge of multilingual alignment highlights the complexities of machine learning, where the same model can yield vastly different outputs based on the language context. For more detail, see VentureBeat AI. The implications of this research go beyond technical aspects; they influence how businesses and governments might implement AI technologies. If a model like Claude can provide different ethical guidelines depending on the language, organizations need to take these variations into account when integrating AI. Anthropic's ongoing efforts to improve Claude's multilingual abilities reflect a broader trend in the industry. The goal is to develop AI that not only performs well but is also ethically sound. As AI becomes a part of everyday life, it's crucial to ensure these systems operate fairly across different languages to build trust and acceptance among diverse users. What's Next Looking forward, the AI community faces the challenge of addressing these findings. As Claude continues to evolve, future updates may include features aimed at better aligning its values across languages, possibly resolving the inconsistencies we've seen. This could involve training the model on diverse datasets that capture a wide array of cultural values and ethical frameworks. Plus, as Anthropic refines Claude's capabilities, the company may pursue partnerships with academic institutions for further research into multilingual alignment. Such collaborations could be key in developing systematic approaches to tackle the risks associated with language-dependent ethical variations. As the tech landscape evolves, organizations using AI must stay aware of the ethical aspects of their tools. This means actively engaging with findings from Anthropic to ensure they're using AI responsibly in multilingual contexts. FAQs Why does Claude exhibit different values in different languages? Claude's values can shift based on the language because of the model's training data and the cultural nuances tied to each language. How does this affect the deployment of AI across different regions? Organizations need to be mindful of these value differences when deploying AI since the ethical implications can lead to misunderstandings or misapplications in various cultural contexts. What steps can developers take to mitigate these inconsistencies? Developers should focus on training AI models with diverse datasets that showcase a range of cultural values, helping create a more balanced ethical framework. Is this issue unique to Claude? While Claude's situation highlights the challenge, similar issues have been observed in other AI models, pointing to a wider need for better multilingual alignment across the industry. How significant is the context window of Claude 3 Opus? The 200,000 token context window enables Claude 3 Opus to manage complex queries and retain information over lengthy interactions, improving its overall usability. Stay tuned for more updates from Anthropic.
Cathie Wood Buys The Dip In SpaceX Stock Despite AI Bubble Warning By Experts
Sponsored This page may contain affiliate links. If you sign up through these links, we may earn a commission at no additional cost to you. This does not influence our editorial reviews or rankings. Even as investor sentiment was dampened by concerns about an AI bubble, Cathie Wood's ARK Invest kept accumulating a pile of Elon Musk's SpaceX stock. The latest acquisition came as the SPCX share price continued dropping, nearing the IPO price of $135. Cathie Wood Continues Buying SpaceX Stock On Monday, July 13, the SpaceX stock closed at $139.14, 4.24% lower in the intraday session. During the day, ARK Invest purchased 130,241 shares of SPCX into three exchange-traded funds, according to its daily trading disclosures. Despite the recent weakness, this purchase represents Cathie Wood's bullish take on the SpaceX stock. ARK Invest raked in SPCX shares via its ARK Innovation ETF (ARKK), the ARK Autonomous Technology & Robotics ETF (ARKQ) and the ARK Next Generation Internet ETF (ARKW). The total investment came around $21.3 million. The latest drop has brought the stock price back down below the reported crucial level of $150. Now, $145 serves as a major resistance level that earlier acted as a crucial support level, per CoinGape analysis. If the share price continues to decline, it risks falling below the IPO price of $135. However, the analysis went on to highlight that the shares had rallied after ARK Invest's buying spree last week. At the time, Cathie Wood's ARK bought $52 million in SpaceX stock. Still, the MACD indicator had taken a negative turn, indicating bearish momentum was still in place. This could keep the stock price from getting back to $150, the analysis added. The AI Bubble Warning In Focus The surge in Cathie Wood's purchasing activity for SpaceX coincides with a draft report from the U.S. Department of the Treasury. That report has raised concerns that the rapid development of AI could be a threat to the economy comparable to that of the dot-com bubble. Researchers in the field of careers at the University of Texas, Austin (NOTUS) have determined that the companies providing artificial intelligence are much more interwoven with the U.S. economy as compared to internet companies 25 years ago. The report states that the impact could radiate beyond tech stocks. They warned that it could spread to private credit markets, cloud service providers and semiconductors as well as utilities and firms financing large-scale data center projects if the AI sector goes into a major downturn. The analysts stopped short of forecasting an imminent crash. Rather, they laid out a bearish scenario in case AI companies do not achieve the level of productivity and profitability that is expected. Here, they warned investment growth could slow, investor confidence could be dampened, and economic growth could be reduced. Supply chain disruptions, geopolitical tensions, electricity shortages and funding limitations for data centre infrastructure were also identified as risks, the report noted. What Do Experts Say? Meanwhile, market watchers are speculating on whether the AI craze has gone too far. In a recent Substack post, Bernstein and Cummings say that the gains of the leading AI stocks suggests a bubble that is "still inflating." The analysts also said, "The aggressive push into AI is leaving these tech giants with significantly less cash." Moreover, they added that technology-based investments have reached nearly 5% of the U.S. GDP, surpassing the dot-com era. On the other hand, BlackRock analyst Rick Rieder stated that the asset management firm would be cutting its stake in companies focused on artificial intelligence. Instead, they are looking to move towards those that will benefit indirectly from the AI boom. One of those beneficiaries is Bitcoin miner TeraWulf that has signed a 20-year contract with Anthropic to host one of its data centers.
Elon Musk Says SpaceX Will Be Worth More Than the Entire Earth Because Musk Will Harvest the Sun
The number he's chasing is Earth's entire GDP, roughly $100 trillion. His pitch to investors: capture a sliver of the Sun's energy from space, and one company clears the whole planet's economy. Musk floated the claim in a tweet defending SpaceX's recent AI compute deal with Anthropic, after a supporter argued the deal handed too much money to a rival. His reply: "SpaceX will be worth more than the rest of Earth if we accomplish our goals." The defense has teeth. Anthropic is paying SpaceX $1.25 billion a month through 2029 for full access to the Colossus 1 data center in Memphis (300-plus megawatts, hundreds of thousands of NVIDIA GPUs), and Google signed a similar deal at $920 million a month. Combined, SpaceX's AI compute contracts are reportedly running at a $26 billion-plus annual rate, and Anthropic has already expressed interest in SpaceX building multiple gigawatts of compute capacity in space. The logic runs through the Kardashev scale, a measure of how much energy a civilization can channel. Musk figures space-based solar power could tap roughly 100,000 times more energy than Earth while using less than a millionth of what the Sun puts out, enough, he says, to eclipse global GDP. The energy math isn't crazy on its face. Earth currently runs on about 2 x 10¹³ watts of average power. Multiply that by 100,000 and you're at 2 x 10¹⁸ watts, which is roughly one two-hundred-millionth of the Sun's total output of 3.8 x 10²⁶ watts. No Dyson sphere required, just a swarm of orbital solar satellites beaming power down, assuming launch costs keep collapsing. Concepts already on the drawing board target single installations delivering 2 gigawatts continuously to Earth, one array doing the work of several ground plants with almost no weather or night losses. The economic leap follows the same logic. Every previous jump in cheap energy (coal, then oil and electricity) produced a non-linear jump in GDP per person, because new energy regimes don't just scale existing activity, they unlock industries that were previously uneconomic (synthetic fuels, desalination at scale, orbital manufacturing, AI clusters that couldn't be powered before). Stack a 100,000x energy unlock on top of AI removing cognitive bottlenecks and robots like Optimus removing labor bottlenecks, and the ceiling on annual output stops being set by physics. That's why the orbital compute angle matters: AI training is already slamming into terrestrial power and cooling walls, and space offers near-constant sunlight and passive vacuum cooling. The bet is that SpaceX owns the infrastructure layer (Starship for cost-to-orbit, Starlink for comms, orbital compute deals like the Anthropic one) that any of that gets built on. He's made versions of this pitch before. In January he said "space-based industries will vastly exceed the value of all of Earth," and he's claimed Tesla's Optimus robot will raise "Earth GDP by an order of magnitude." The company those promises are attached to lost $5 billion last year. SpaceX shares rallied after a blockbuster IPO last month, then slid, with investors still wary of a multitrillion-dollar valuation. And on July 10, China recovered the first stage of a Long March 10B rocket via net capture on a ship, its first orbital-class reusable booster landing, closing the reusability gap that made SpaceX's economics work in the first place. Thanks for spending time with Thought Catalog. Connect with us on Facebook and browse the rest of our website.
Cursor poaches designer behind Anthropic's Claude
The hire highlights the growing competition for AI product talent, as companies increasingly compete on user experience -- not just model performance. Jenny Wen, the designer involved in creating Anthropic's Claude and the general-purpose AI agent Cowork, has joined Cursor as its new head of design. She will be leveraging her experience at one of the most established AI labs at a rapidly growing company that aims to develop more than coding tools. Wen disclosed the news on X on July 13, after leaving Anthropic to take the position at Cursor, where she would be able to lead "a team that cares so deeply about craft, quality, and building great tools." She chose to make this decision at a time when Cursor is developing its own AI agent aimed at helping in the workplace, which puts it in competition with the product that Wen was involved in at Anthropic. From Claude to Cursor During her time at Anthropic, Wen had the task of designing both Claude and Cowork, the latter being the name of the AI agent that was designed to handle complex workplace tasks. According to Wen's portfolio, in 2025 Wen was heavily involved in directing the roadmap of Cowork as well as contributing to the development of the product post-launch. Before Anthropic, Wen worked at Figma as its Director of Design, where she launched FigJam, and previously served as the design leader at Dropbox, Square, and Shopify. The uniqueness of Wen's experience lies in the integration of product strategy and product creation. With more AI companies competing to develop better models, it is worth remembering that the skill to create desirable products remains just as crucial as the technical part. Initially recognized as an AI coding assistant, Cursor now aspires to pursue more ambitious goals. According to The Information, the company is now working on the creation of an AI assistant called Sand, which can manage e-mails and spreadsheets and help with engineering tasks. If released successfully, it will compete with Cowork by Anthropic and the recently introduced ChatGPT Work by OpenAI. Cursor has not made any announcements regarding the project. Additionally, the company is gearing up for its next growth stage. SpaceX declared in June that it has intentions to take over Anysphere, the parent company of Cursor, in an all-stock transaction worth $60 billion expected to close later this year. The agreement will give Cursor more resources, thus enabling it to widen its scope beyond just developing software. Why this matters for Anthropic? It is interesting to note that the timing takes place at a time when Anthropic has expanded the capabilities and accessibility of Cowork. The company has updated the service for use on both mobile and the web, as well as added the cloud execution functionality, which enables tasks to continue even if the user has disconnected. Anthropic even states that, out of its analysis of about 1.2 million Cowork sessions, knowledge-work tasks make up almost half of all usage, suggesting the product is increasingly being used for more than software development. While Wen's leaving may not indicate difficult times ahead for Anthropic, it points to the intense competition among AI firms for talented people like product leaders. Now that emerging AI technologies are growing in sophistication, the challenge is not just about developing better technologies, but also about creating software that easily fits into the routines of end-users. This is where experienced design leaders can make a big difference. Creating an effective workplace AI agent involves more than just technical performance. In addition to this, it requires proper design of workflows, simple interfaces, and a lot of trust for users to delegate important tasks. Those are qualities which, if developed, can give a strong competitive advantage. A broader shift in the AI talent market This step by Wen is indicative of a bigger trend emerging within the competitive environment surrounding AI enterprises. Whereas back in the days of early generative AI, the battle for talent was mainly carried out between researchers and infrastructure engineers, now there is competition among businesses to secure experts who already have experience working on creating and deploying AI products. These experts will be able to share valuable information about consumer behavior and preferences, as well as features that actually work. In the case of Cursor, the recruitment of the designer of one of the first AI workplace assistants will allow the company to go beyond programming in terms of further steps in its development. In the case of Anthropic, the hiring shows how the market has developed, with competition, among other things, both for skilled workers and for the designers behind those who turn advanced AI into products.
Why India Became Anthropic's Biggest AI Bet As Claude Gets Local Pricing? - The Logical Indian
Anthropic launches India-specific Claude pricing as AI competition intensifies. India now contributes 5.8% of global Claude.ai usage, making it Anthropic's second-largest market by usage share after the United States. But the company's decision to introduce India-specific rupee pricing is not only about making subscriptions easier to buy. It reflects a larger strategy to convert strong developer adoption into a sustainable business opportunity in one of the world's fastest-growing technology markets. Anthropic's move comes as global AI companies compete for India's developers, enterprises and startups, where adoption is increasingly shifting from experimentation to workplace applications. India Emerges As AI Hub Anthropic's data shows that India's Claude usage is heavily concentrated around technical work. Nearly half of Claude activity in the country involves computer and mathematical tasks, including software development, system modernisation and production software deployment. This usage pattern sets India apart from markets where AI adoption is driven primarily by consumer productivity tools. For Anthropic, Indian users represent a segment that is already integrating AI into professional workflows. However, total usage does not necessarily mean widespread adoption across the population. While India ranks second globally by overall Claude usage share, Anthropic's data shows that the country ranks much lower on Claude usage per working-age population. This suggests India's position is being driven largely by its scale, technology workforce and developer ecosystem rather than universal AI adoption. Rupee Pricing Targets Growth Anthropic's India pricing rollout addresses one of the practical barriers faced by local users: paying for a global subscription priced in foreign currency. The company has introduced India-specific pricing for Claude plans, with Claude Pro available at ₹2,000 per month when billed annually. Claude Max starts at ₹11,999 per month, while Claude Team plans begin at ₹2,399 per seat per month. The move simplifies pricing visibility for Indian customers by removing direct dollar conversion calculations. However, it does not represent a complete shift towards India-specific payment infrastructure. Reports indicate that payment options remain different from some India-focused digital platforms, with UPI availability not being part of the initial rollout. Anthropic's India pricing also does not appear to follow a low-cost market strategy. The India-listed Claude Pro annual price is higher than the US base price of $17 per month before taxes. However, comparisons are affected by currency exchange rates and different tax structures across markets. AI Companies Chase India Anthropic's localisation push comes amid intensifying competition among global AI companies looking to establish their position in India. The country offers a combination of a large software workforce, a rapidly expanding startup ecosystem and increasing enterprise interest in artificial intelligence. For AI firms, India represents both a large user base and a potential source of developers building applications around AI models. The challenge for companies like Anthropic is moving beyond free experimentation. While millions of users interact with AI tools globally, converting regular users into paying subscribers remains a major commercial challenge across the industry. In India, that challenge is closely linked to pricing, accessibility and whether businesses find enough value in integrating AI tools into everyday operations. Building Long-Term Presence The pricing announcement follows Anthropic's broader expansion strategy in India. The company opened its Bengaluru office in 2026 and appointed Irina Ghose, former Microsoft India Managing Director, to lead its India operations. The Bengaluru presence indicates that Anthropic views India as more than a market for selling subscriptions. The company is also looking to build partnerships and strengthen its enterprise presence. As competition increases among AI providers, success in India may depend on more than model capability. Pricing structures, local support, partnerships and developer adoption will play an important role in determining which AI platforms become deeply embedded in workplaces. For Anthropic, India's importance is already visible through usage patterns. The next challenge is converting that engagement into long-term customer relationships and commercial growth. News in FAQ 1: Why has Anthropic introduced rupee pricing for Claude AI in India? Anthropic has introduced rupee pricing to make Claude subscriptions easier for Indian users by displaying prices in local currency. The move simplifies billing, reduces currency conversion friction and reflects India's growing importance as one of Claude's largest usage markets. 2: How much does a Claude AI subscription cost in India? Claude Pro costs ₹2,000 per month with annual billing, Claude Max starts at ₹11,999 per month, and Claude Team plans begin at ₹2,399 per seat per month. Prices may vary depending on the selected plan and billing cycle. 3: Why is India important for Anthropic? India accounts for 5.8% of global Claude.ai usage, making it Anthropic's second-largest market by usage share after the United States. The company also sees strong demand from developers, startups and enterprises using AI for technical and business applications. 4: Is Claude AI cheaper in India than in the US? Not necessarily. While Claude is now priced in Indian rupees, the India-listed subscription price is broadly aligned with global pricing and may appear higher than the US base price because of taxes, exchange rates and local pricing structures. 5: What is Claude AI mainly used for in India? According to Anthropic, nearly half of Claude usage in India involves computer and mathematical tasks. These include software development, coding assistance, system modernisation, debugging and deploying production software across businesses and technology teams.
Anthropic introduces Rupee-based pricing for Claude AI amid big India push
India is home to one of largest Claude user communities, and the company continues to explore ways to make its products more accessible in the region Anthropic has unveiled Rupee-denominated pricing for Claude AI assistant, simplifying payments for customers in India, a market that boasts of one of the largest Claude user communities. Claude Pro plan (more usage, access to Claude Code and Cowork) is available at ₹2,000 per month, and Max (with higher limits and priority access at high traffic times) is being offered from ₹11,999 per month. The Rupee-based pricing for Claude AI subscription can be seen on its website and mobile apps; the program currently supports card payments. Anthropic is testing local pricing in Indian rupees for new consumer subscriptions of Claude Pro, Max, and Team plans, as part of a broader push to offer local pricing across markets, people familiar with the development said. India is home to one of largest Claude user communities, and the company continues to explore ways to make its products more accessible in the region. Also Read Crude oil tops $87: Will it derail the big FPI comeback on D-Street? Freshers vs experienced professionals: Who is winning the AI hiring race? LTM expects AI revenue to outpace traditional services: CEO Venu Lambu Economists, tech leaders urge governments to prepare for AI job disruption Rupee falls 48 paise to 96.16 against US dollar during early trade The latest move comes as global AI companies like Anthropic and OpenAI are doubling down on India market. In February this year, Anthropic opened its office in Bengaluru and announced various partnerships across India, describing the market as the second-largest for its Claude.ai and home to a developer community doing "some of the most technically intense AI work". The Bengaluru office is Anthropic's second in Asia after Tokyo. At that time, an Anthropic statement had said that the company's run-rate revenue in India doubled since the expansion announcement in October 2025, and the range of organisations building on Claude reflected the broad distribution of growth from large enterprises to digital-native companies to startups shipping their first products. Anthropic has been drawing attention over its blockbuster funding, soaring valuations and the growing interest in its advanced AI model 'Mythos'. The model has demonstrated an ability to identify previously undisclosed vulnerabilities in widely used software codebases, making it a potent tool for defensive cybersecurity, whie at the same time raising concerns about its misuse. On Monday, IT Secretary S Krishnan had said that getting access to Mythos and other advanced frontier models is high on the government's priority, and while talks and processes for the same are on, alternate AI models are also being leveraged for now to track and fix vulnerabilities. OpenAI too has stepped up its India focus. The Sam Altman-led AI company recently appointed Prabhjeet Singh as Managing Director for India, one of the company's top-priority markets. More From This Section Apple, Samsung pull ahead as memory crisis hits smartphone market in Q2 Apple releases iOS 27 public beta for eligible iPhones: Check what's new Anthropic introduces India pricing for Claude subscription plans: Report Cert-In operationalises AI war room to work on frontier AI models Apple's lawsuit threatens to disrupt OpenAI's bid to rival the iPhone
Canada Regulator Warns Banks About Cyber Risks of Anthropic's Claude Mythos
Canada's federal banking regulator warned the country's largest financial institutions about the risks of Anthropic's Claude Mythos and other advanced AI models, saying the new technology could increase cyber threats and reduce the time institutions have to identify and fix vulnerabilities, according to an email sent in April. The regulator, the Office of the Superintendent of Financial Institutions, sent the email to chief technology officers, chief information security officers, and chief risk officers across the financial industry, including the big banks and insurers, according to documents Reuters obtained through an access-to-information request. Regulators globally are trying to assess cybersecurity risks such as Anthropic's frontier AI model Mythos. Cybersecurity experts say Mythos, an AI model described as extremely capable at finding and exploiting cybersecurity vulnerabilities, poses significant challenges to the banking industry and its legacy technology systems. "Advanced artificial intelligence models, such as Anthropic Claude Mythos, significantly compress the timeframe for effective risk mitigation," OSFI said in the email. "Accordingly, this bulletin is grounded in our existing guidance and outlines sound practices that institutions can adopt to enhance the speed and effectiveness of risk identification, mitigation and response." Additional contents of the email were redacted due to some sections of the Access to Information Act. An acknowledgment of the risks of Mythos from OSFI could ensure Canadian banks, insurers and other regulated institutions invest in technology to protect clients from cyber risks. After Reuters sent questions to OSFI last week, the regulator on Monday posted a public bulletin on generative and agentic artificial intelligence online. "OSFI takes a technology‑neutral, risk‑focused approach to emerging technologies, including advanced artificial intelligence models such as Mythos. Our focus is not the technology itself, but how federally regulated financial institutions govern and manage the risks associated with its use," the regulator said in an emailed response to Reuters questions. In early April, Canadian bank executives met with regulators to discuss the risks posed by Mythos shortly after U.S. Treasury Secretary Scott Bessent and then-Federal Reserve Chair Jerome Powell convened an urgent meeting with bank CEOs to warn of cyber risks posed by Anthropic's latest artificial intelligence model. OSFI sent the email to company executives on April 29. Rapidly Changing Landscape OSFI is responsible for regulating and maintaining the stability of Canada's financial sector, from banks to pension funds, and identifying risks emerging from foreign interference, geopolitics and new technology. The cyber capabilities of some frontier AI systems are considered so powerful that access has been restricted, with euro zone banks currently excluded from Mythos. Anthropic has also had a tumultuous relationship with the U.S. government. A judge blocked its initial blacklisting by the Pentagon in March, and the conflict has eased following the private release of Anthropic's Mythos. Three of Canada's big six banks - Royal Bank of Canada, TD Bank and BMO - have outlined a plan to earn millions from their investments in AI as the banks moved from experimental AI projects to applying them in chatbots, building internal tools and lowering their reliance on third-party tools. Bank of Nova Scotia, CIBC and National Bank have also disclosed several AI initiatives. The Canadian government has said it has access to Anthropic's Project Glasswing, which allows companies to have access to Mythos. It is not clear which, if any, banks in Canada are using it. Some banks deferred comments to the Canadian Bankers Association, which said banks have invested heavily to protect the financial system and are complying with robust requirements from OSFI on cyber risk management and incident reporting. In an interview in June, RBC's AI Group Head Bruce Ross said Mythos underscored a shift in the cyberattack landscape, making it imperative for organizations to respond rapidly since attack methods can emerge as soon as new vulnerabilities are identified. "The way we're (the industry) dealing with it is, building our own AI defenses... we'll continue to do that," Ross said. (Reporting by Nivedita Balu in Toronto; editing by Caroline Stauffer and Deepa Babington)
Anthropic Wrongly Bills a Free-Tier User $16.6 Million: What Happened?
An audit by Vaudit found $1.7 million in overcharges across $34 million in AI invoices. Anthropic confirmed a massive billing error after its system tried to charge a South Korean free-tier user $16.6 million, despite zero API usage and no registered card. The phantom invoices blocked his credit card and triggered a four-day support ordeal before the company responded. Inside Anthropic's $16.6 Million Phantom Invoice A phantom invoice is a payment demand generated by a billing system without any underlying transaction or usage. That is exactly what a Korean developer, known as remy_notes, received from Anthropic in early July. The story quickly went viral after he shared the screenshots on Threads. The first invoice arrived on July 7, totaling $1,669,875. Less than 24 hours later, a second one multiplied the figure almost tenfold to $16,627,739. The escalation occurred without any activity being recorded on the accoun Follow us on X to get the latest news as it happens. The developer initially suspected phishing. However, the emails came from Anthropic's official domain and used the company's legitimate Stripe infrastructure. Furthermore, his dashboard showed zero API usage, no billable keys, and no payment method on file. He audited his own automation scripts and AI agents, finding nothing that explained the charges. His bank declined two overseas charge attempts on July 8 because the amounts exceeded the card's per-transaction limits. The repeated attempts still resulted in his primary credit card being blocked. What Caused the Billing Error, According to Anthropic Anthropic confirmed the mistake on July 12 and assured that no money was collected. The company attributed the invoices to an incorrect automatic credit reload setting, which it disabled as a precaution. It also clarified that the incident did not involve unauthorized access. Anthropic has not explained how the reload value reached such an extreme figure. The resolution, however, was far from smooth. The user spent four days and around 18 emails seeking written confirmation that the invoices were void. Meanwhile, automated replies kept arriving instead of human responses. He eventually asked the company to respond by Monday afternoon, Korean time. The case also exposed a wider industry problem. Audit startup Vaudit reviewed $34 million in AI invoices from 60 enterprise clients and found roughly $1.7 million in overcharges. That represents an error rate of about 5%, much of it tied to AI billing systems. Clients affected by those errors included major corporations like Panasonic, HP, and Honda. For users, the lesson is simple. Reviewing billing dashboards, monitoring card notifications, and reporting unusual charges quickly can prevent a glitch from becoming a financial nightmare.
:quality(82)/ba125d6198354767b4378a38de405963)