Anthropic briefly tested moving Claude Code to $100/month tier, then reversed the change

On April 22, 2026, Anthropic quietly updated its pricing page to restrict Claude Code—previously bundled with the $20/month Pro subscription tier—to Max plans costing $100 or $200 per month. The change appeared only on claude.com/pricing and not on support documentation, creating inconsistency that amplified confusion once users noticed. Internet Archive snapshots captured both the updated and reverted states within a 24-hour window.

Amol Avasare, Anthropic's Head of Growth, disclosed on Twitter that the update was part of a limited experiment affecting approximately 2% of new 'prosumer' signups, with no impact on existing Pro or Max subscribers. A subsequent clarification acknowledged that updating public landing pages and documentation for a narrow A/B test was an operational error; Anthropic subsequently reverted both, while stating the underlying experiment would continue invisibly.

The public-facing reversal occurred within hours of the initial update—roughly the time required for discussion threads to surface on Reddit, Hacker News, and social media. No formal announcement or statement of intent followed the reversion, leaving ambiguity about whether the test remains active, how results would inform future pricing decisions, or what threshold of user price resistance might halt the experiment.

The episode coincided with competitive pressure from OpenAI, whose Codex product engineering lead publicly reaffirmed Codex's availability across free and $20 tiers, characterizing Anthropic's handling as a missed opportunity to reinforce trust through transparent communication.

For educators, developers building on Claude Code, and regions where subscription costs represent material expense relative to income, the test—and its reversal without clarification—introduced strategic uncertainty about the product's accessibility roadmap, affecting longer-term adoption and build decisions.