Auto mode’s token overhead stops counting against your limits
All about: Claude Code versions and changes →
In auto mode, a classifier inspects every tool call, and that check burns a small number of extra tokens. As of today, those tokens no longer count toward usage limits on Pro, Max, and Team.
Enterprise and API users get the same treatment once auto mode becomes the default for them. On Pro, Max, and Team, it switches on by default August 14.
Up next
Managed Agents can now pin which region runs the compute
Claude Managed Agents can now pin where inference runs: model.inference_geo set to “global” routes requests to whatever capacity is free at standard pricing, while “us” keeps compute in region at a 1.1x rate. Anthropic’s docs say it’s enforced at every agent step.
Related stories
- Claude Code can now resume itself after a limit reset
- Auto mode let Claude Code run downloaded code in 6 of 10 tests
- Claude Code turns auto mode on by default August 14
- 44 KB of Claude Code’s classifier prompt never gets printed
- Claude Code cuts subagent update cost 99%, session use 5%
- Claude Max, Ultra and Team plans add a monthly API credit
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.