Claude over AI Code? Anthropic in fresh drama as users say access to Code made more expensive

Internet erupted in outrage after Anthropic seemingly removed Claude Code access from its Pro plan on its website. (Representational image made with AI)

If you have spent some time on the internet lately, Anthropic is a name you likely would've heard. The AI startup that is after your job, or at least is creating models that can do your job better than you. Within five years, Anthropic has reached a valuation of $350 billion. And the crown jewel for Anthropic is Claude Code - an AI coding tool that can write hundreds, if not thousands of lines of code by itself, perhaps even better than you. But Claude Code users have started getting frustrated. They say Anthropic is purposely trying to make Claude Code more expensive to use.

On X, some users noticed that Anthropic had quietly updated its paid plans. The Claude Pro plan, which costs $20 (roughly Rs 1,876), no longer gave you access to Claude Code, something that was previously available.

This meant that anyone who wanted to use Claude Code would need to shell out for the Claude Max plan that costs $100 (roughly Rs 9,380). A five-fold increase to the tool they may have been using for months.

And there was outrage. One user called it "borderline suicidal." The user wrote, "Removing claude code entirely is such an idiotic move when everything you are known for is coding." Another person added, "Anthropic stripping claude code from the pro plan is a massive betrayal."

Anthropic's chief rival, OpenAI, found this to be the perfect opportunity. The head of OpenAI's Codex coding tool, Thibault 'Tibo' Sottiaux wrote, "I don't know what they are doing over there, but Codex will continue to be available both in the FREE and PLUS ($20) plans." CEO Sam Altman swiftly replied, "We want you to have a lot of AI!" And Anthropic, led by Altman's nemesis, Dario Amodei, was now on the back foot.

Following the outrage, Anthropic went into damage control mode. Amol Avasare, a staff member, wrote that it was actually a test that was only for new users. He posted on X, "For clarity, we're running a small test on ~2 per cent of new prosumer signups." Avasare insisted that "Existing Pro and Max subscribers aren't affected."

But the damage was done. People were asking why this change, which was only for 2 per cent of new users, was visible to everyone on the website. In a separate post, Avasare simply stated that the company had now gone back to the old pricing on the website. He wrote, "This was understandably confusing for the 98 per cent of folks not part of the experiment, and we've reverted both the landing page and docs changes."

Sam Altman, who was perhaps having too much fun looking at Anthropic's troubles, replied to this post with a classic Gen Z meme. He said, "ok boomer."

This is not the first time Anthropic has found itself under scrutiny by users. Last month, users complained that they were running out of their Claude token limits much faster than before. Tokens act as a unit of measurement for AI models. The more intensive the work, the more tokens you consume. Do note that Anthropic doesn't actually tell you the number of tokens you get in your Claude plan.

So some were worried that Anthropic may be reducing the token limits while charging the same subscription fee.

After investigating the issue, Anthropic said that no one was billed unfairly, and the company advised the Claude Pro users to switch to the less powerful Sonnet models instead of the more advanced Claude Opus. In a way, users were being told to use the less powerful models even though they were still paying the $20 subscription fee.

And it appears that the internet is noticing a pattern now. One user saw the 2 per cent test as a "classic shrinkflation playbook." Shrinkflation is when you pay the same price but get a reduced quality of a product. The user noted that in the past few weeks, Anthropic had reduced functionality for Claude.

They explained, "Last month your $20 Pro plan powered OpenClaw, Cursor workflows, and anything else you plugged into it." The user added, "This month those same tools now pull from a separate plan. Some power users had to go from $200/month to $1,000+ overnight."

It may seem odd to see a company that is rolling out new updates almost every day or so to be in this situation. After all, the internet is full of memes that go along the lines of, "I wake up and there's a new Claude update." But Anthropic likely isn't doing this out of greed, it is potentially struggling with compute to run its AI models.

Large Language Models (LLMs) like Claude or ChatGPT run on data centres.Thanks to the surge in demand for AI, companies are finding it difficult to keep up. Reports indicate that Anthropic has not been as aggressive as OpenAI when it comes to buying up computing power. While the company has signed a new deal with Amazon to use its compute infrastructure, it seems that Claude's demand might be too high for Anthropic to keep up unless it gradually increases prices.