Anthropic Reverts A/B Test Removing Claude Code from Pro Subscription Plan

This test reflects broader trends in AI service pricing and user engagement, impacting how companies structure their offerings amid rising compute demands.

• On April 21, 2026, Anthropic conducted a limited A/B test removing Claude Code from its $20 Pro plan for about 2% of new subscribers.
• User backlash erupted on platforms like Reddit and X, leading Anthropic to revert the changes within a day.
• Amol Avasare, Anthropic's Head of Growth, clarified that the test was a response to evolving usage patterns and compute constraints.

• Claude Code is a tool within Anthropic's AI suite that has gained popularity due to its integration with other services, leading to increased resource demands.
• Usage patterns shifted from short interactions to longer, more complex workflows, necessitating a reevaluation of pricing structures.
• The $100+ Max plan was initially designed for heavy chat users but has absorbed features to accommodate high-demand users, reflecting the need for tier restructuring.

On April 21, 2026, Anthropic initiated a pricing experiment that aimed to address the surging compute demands associated with its Claude Code tool. This tool, integral to the Claude AI suite, had seen explosive adoption following its integration with Opus 4 and Cowork, shifting user engagement from brief chat sessions to prolonged, resource-intensive workflows. As a result, the company faced increasing pressure to manage compute costs effectively.

The decision to remove Claude Code from the $20 Pro plan for approximately 2% of new subscribers was a strategic move to test the waters of pricing adjustments amid these evolving usage patterns. However, the immediate backlash from users was swift and vocal, with many expressing frustration over the perceived devaluation of their subscriptions. Developers, in particular, voiced concerns that the removal rendered the Pro plan "useless," leading to demands for refunds and clarifications.

Anthropic's quick retraction of the changes underscores the delicate balance companies must maintain between pricing strategies and user satisfaction. The backlash was not just a reaction to the removal itself but also highlighted broader concerns about the sustainability of AI pricing models in the face of rising compute costs. As users increasingly rely on AI tools for complex tasks, the pressure on companies like Anthropic to provide value while managing operational costs intensifies.

The incident also reflects a growing trend in the AI sector where companies are experimenting with tiered pricing structures to accommodate varying levels of usage. The $100+ Max plan, initially aimed at heavy chat users, has absorbed high-demand features, indicating a shift in how companies are structuring their offerings to meet user needs. This restructuring is essential as the landscape of AI usage continues to evolve, with longer workflows becoming the norm.

In summary, Anthropic's A/B test serves as a case study in the challenges of pricing AI services amid shifting user engagement patterns. The rapid response to user feedback illustrates the importance of maintaining customer trust while navigating the complexities of operational costs and service offerings.

• Developers: Frustrated by the removal of key features, they demand refunds and express concerns over the value of their subscriptions.
• AI Enthusiasts: Users who rely on Claude Code for complex workflows feel the immediate impact of the changes.
• Market Analysts: Observers of AI pricing models are keenly interested in how this incident reflects broader trends in the industry.

• Future pricing experiments: Keep an eye on Anthropic's upcoming tests and adjustments, as they may signal shifts in how AI services are priced and structured.
• User engagement metrics: Monitor how changes in service offerings affect user retention and engagement, particularly among developers and heavy users.
• Competitive responses: Watch for how competitors like OpenAI Codex react to Anthropic's pricing strategies and user feedback, which could influence market dynamics.