New Anthropic Tool Speeds up AI Agent Development for Enterprises

The product comes amid a flurry of releases from the generative AI vendor, which is locked in a tight battle with rival OpenAI.

Anthropic released a new tool designed to make it easier and quicker for businesses to build and deploy AI agents.

Claude Managed Agents provides developers with out-of-the-box infrastructure that, according to Anthropic, will enable enterprises to launch agents 10 times faster than before.

The system achieves this by taking responsibility for the months of complex development work that is usually required before a production agent is shipped, eliminating the need for businesses to do this themselves, according to the vendor.

A blog post about the tool said this pre-production work can include "sandboxed code execution, checkpointing, credential management, scoped permissions, and end-to-end tracing".

Now, though, none of that is required, Anthropic said. Instead, customers define an agent's tasks, tools and guardrails and run it on Anthropic's infrastructure.

The automated spins up a container for each agent, which can decide when to call tools, manage context and recover from errors.

Other claimed benefits include long-running sessions that can run for hours, and coordination with other agents, enabling complex work to be done in parallel, although this latter feature is currently only available in research preview.

Also in research preview is a feature that allows Claude to continually refine and self-evaluate response quality until it meets the success criteria defined by the user.

Managed Agents is currently available in public beta on the Claude platform, and Anthropic lists an array of big-name users, including Notion, Rakuten, Asana and others.

Rakuten, for example, is using agents across its product, sales, marketing, finance and HR teams, with each deployed within a week. Removing the need to develop infrastructure has allowed the rewards platform vendor to "focus entirely on democratizing innovation across the company," Yusuke Kaji, general manager, AI, at Rakuten, said in a statement.

Pricing for the new tool is based on consumption, with users paying for the models' token use, as per Anthropic's standard API costs, plus an extra $0.08 per session-hour for active runtime.

Earlier this week, Anthropic CFO Krishna Rao revealed that the company's run-rate revenue had now surpassed $30 billion -- more than tripling from $9 billion at the end of 2025 -- and that the launch of Managed Agents appears timed to capitalize on the rapidly evolving enterprise market and accelerate that growth further.