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
@bcherny
-
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
-
Rate limits adjusted for improved xhigh tier usage
By
–
We’ve tuned rate limits to give you the same amount of usage with xhigh
-
Adaptive thinking budgets replace fixed AI reasoning limits
By
–
Yes. We no longer support fixed thinking budgets, and find that adaptive thinking works much better. Tune it with /effort
-
Claude 4.7 xhigh effort level balances reasoning latency token usage
By
–
In Claude Code the default effort is now xhigh, a new level between high and max giving finer control over the reasoning/latency tradeoff. 4.7 thinks more, so token use runs higher than 4.6. Manage it with effort, task budgets, or prompting for brevity.
-
Claude Auto Mode: Autonomous Task Execution Without Permission Prompts
By
–
We're also rolling out auto mode for Max users. This means no more permission prompts — give Claude a task, let it run, come back to verified work.
-
Claude 4.7 Enables Complex Multi-File Code Changes in Single Prompt
By
–
With 4.7 you can push a lot further with one prompt. That means multi-file changes, ambiguous debugging, code review across a whole service. The stuff you used to break into small chunks because the model would drift.
-
Opus 4.7 Arrives in Claude Code with Enhanced Capabilities
By
–
Opus 4.7 is in Claude Code today. It's more agentic, more precise, and a lot better at long-running work. It carries context across sessions and handles ambiguity much better.
-
Report Opus bugs via feedback command for debugging
By
–
Would you mind running /feedback and sharing the id here next time you see Opus do something weird? Would be very helpful for debugging