Boris Cherny
@bcherny
Opus 4.7 uses more thinking tokens, so we've increased rate limits for all subscribers to make up for it. Enjoy!
474retweets11klikes
View on XAnthropic Increases Claude Subscriber Rate Limits to Offset Opus 4.7 Thinking
· Updated
Anthropic increased rate limits for all Claude Pro and Team subscribers to accommodate the higher token usage of
Opus 4.7. This model utilizes extended thinking (internal reasoning tokens used to solve complex problems) and a new tokenizer. These changes previously caused users to hit usage caps faster.As models shift toward test-time compute (allocating more processing power during inference to improve accuracy), the cost of "thinking" often falls on the user's quota. By raising the ceiling, Anthropic is decoupling improved reasoning performance from subscription value. This prevents the more capable model from feeling like a downgrade in capacity.
The increased limits are active for all subscribers using Opus 4.7. If you are not seeing the higher capacity, ensure you are running the latest version of Claude Code. This adjustment allows you to maintain a high-velocity workflow without being penalized for the model's more thorough reasoning process.
