Updates

What each bake found, and what's new at FairBake.

9 Nov 2026Bakeoff: Same fixes, 1.8x the bill, depending on the harness
Running 3 Claude models in Claude Code instead of pi cost 1.8x as much on 10 real tasks. The harness changed the bill, not the fixes.
28 Oct 2026Bakeoff: More thinking isn’t (always) better
Turning coding agents up to high effort roughly doubled how much code they changed, and only 2 of 5 models solved more tasks with it. Fable 5.1's default was its worst setting.
22 Sep 2026The same bug fix costs 0.4¢ or $2, depending on which coding agent you ask
277 runs of Claude Code, Codex, and pi on bug fixes from my own repos: an 80x cost spread, speed sorted by harness, and a sandbox setting that hid failures behind green scores.