Claude Code briefly drops from Pro plan

This is the kind of pricing test that makes sense on a cost sheet and looks messy in public. Anthropic is likely probing how much Pro demand is really driven by Claude Code, but the rollout turned an internal experiment into a trust problem.

• –Agentic coding sessions are expensive, so Claude Code is an obvious candidate for tiering or tighter limits
• –Hiding the change inside a public pricing page breaks the expectation of stable subscription packaging for developers
• –The quick reversion suggests Anthropic was measuring backlash as much as conversion
• –If Claude Code moves upmarket later, Anthropic will need clearer messaging and probably more explicit usage-based pricing
• –This matters competitively because Claude Code is one of Anthropic's strongest developer hooks, and weaker access could push users toward Cursor, Copilot, or OpenAI Codex