We've raised rate limits for all subscribers to make up for the increased token usage. Enjoy!
@bcherny
-
MRCR Evaluation Method Replaced by GraphWalks Approach
By
–
MRCR is a bad eval, we use GraphWalks now. More info here:
-
MRCR Evaluation Method Replaced by GraphWalks Approach
By
–
MRCR is a bad eval, we use GraphWalks now. More info here:
-
Opus 4.7 Adaptive Thinking Uses More Tokens Than 4.6
By
–
Not accurate. Adaptive thinking lets the model decide when to think, which performs better. Opus 4.7 also uses more thinking tokens on average than 4.6, which is why we have increased rate limits for all subscribers to make up for it.
-
Rate Limits Increased for All Subscribers
By
–
We've increased rate limits for all subscribers to make up for it. Happy coding!
-
Debugging Web Search Accidentally Disabled in AI System
By
–
Current best guess is gergley accidentally turned web search off. Debugging with him
-
Claude usage policy and account support guidelines
By
–
We don't ban accounts for using claude -p. Can you dm @trq212 your support ticket?
-
Opus 4.7 Intelligence Improvements and Effective Usage Tips
By
–
Opus 4.7 feels more intelligent, agentic, and precise than 4.6. It took a few days for me to learn how to work with it effectively, to fully take advantage of its new capabilities. Will post a few more tips throughout the day, starting with this blog post:
-
Phasing Out MRCR: Rethinking Long Context Evaluation Methods
By
–
We kept MRCR in the system card for scientific honesty, but we've actually been phasing it out slowly. Two reasons: (1) it's built around stacking distractors to trick the model, which isn't how people actually use long context, and (2) we care more about applied
-
MRCR Phased Out: Focus Shifts to Applied Long-Context
By
–
We kept MRCR in the system card for scientific honesty, but we've actually been phasing it out slowly. Two reasons: (1) it's built around stacking distractors to trick the model, which isn't how people actually use long context, and (2) we care more about applied long-context