2 min read AI-generated

Claude Code Quietly Dropped from the Pro Plan — Developers Are Not Happy

Copy article as Markdown

Anthropic silently removed Claude Code from the $20 Pro plan. The developer community's reaction was swift and unforgiving.

Featured image for "Claude Code Quietly Dropped from the Pro Plan — Developers Are Not Happy"

On Monday evening, sharp-eyed developers noticed something had changed on Claude’s pricing page. Claude Code, previously included in the $20/month Pro plan, now appeared exclusively under the Max plan starting at $100/month. No announcement. No blog post. No changelog entry.

What Actually Happened

Anthropic removed Claude Code access for new Pro subscribers. Existing Pro users are not affected — for now. But the support documentation was already fully rewritten, with the page title changing from “Using Claude Code with your Pro or Max plan” to “Using Claude Code with your Max plan.”

Amol Avasare, Anthropic’s Head of Growth, responded on X, calling it a small test affecting just 2 percent of new “prosumer” signups. His explanation: “Usage has changed a lot and our current plans weren’t built for this.” The company is exploring alternatives to maintain a good experience for all users.

Why Developers Are Furious

The issue isn’t the test itself — it’s how it was communicated. Or rather, how it wasn’t. The support documentation was already fully rewritten to reflect the change, which doesn’t exactly match the “tiny 2% experiment” framing.

George Pu’s tweet flagging the change reached nearly 900K views with over 900 retweets. The comments on Hacker News and Reddit are overwhelmingly negative. Many developers feel they were presented with a fait accompli.

The timing is especially painful for those who recently switched from VS Code or Cursor to Claude Code, drawn in by the affordable Pro plan. Now they’re being told to pay $100/month?

OpenAI Seizes the Moment

OpenAI was quick to point out that Codex remains available on both the free and Plus tiers. The contrast couldn’t be sharper.

My Take

I get the dilemma. Claude Code is expensive to run — every session burns through tokens worth several dollars. The math doesn’t work at $20/month long-term. That’s not surprising. But silently changing the pricing page without any communication? That’s a trust breach Anthropic can ill afford.

If it’s truly just a 2% test, a quick blog post would have been enough: “We’re testing new pricing models for Claude Code. Existing customers aren’t affected.” Instead, they updated the docs and hoped no one would notice. That didn’t work.

Sources: