Source Intelligence
Follow Discord

DisclaimerUnofficial, and not affiliated with Anthropic. Nearly all of this is read straight out of what ships: npm bundles, captured prompts, published docs. Anthropic's own notes go in verbatim, marked as theirs. The rest is my reading, and every entry carries the strings behind it. If one looks wrong, vote it down and say why.

← All of v2.0.26 Home All releases ←olderv2.0.25 v2.0.27newer→
Claude Code v2.0.26

Removal of "Always Thinking" Experiment

What: The experimental "always thinking" feature has been completely removed

Details:

  • Removed automatic ultra-thinking token allocation based on experiment flag
  • Simplified message rendering by removing experiment-specific UI logic
  • Removed from both backend logic and frontend rendering components
  • Evidence: Removed logic from function L_() at line 296620 (was Uk() at line 296382 in v2.0.25), simplified rendering in function n_() at line 323185 (was mk() at line 324226 in v2.0.25, 45.5% similarity)

See this entry in the whole of v2.0.26 →