Reasoning effort study: top effort costs up to 42x more on tasks low effort already solves
zainhas · x · 2026-10-29
- @KaiserWhoLearns ran a systematic test of models × reasoning effort levels × benchmarks, shedding light on the notoriously opaque topic of reasoning token spend in coding agents.
- Key finding: on tasks that low effort can already solve, top effort costs up to 42× more with no added benefit.
- The catch: you can't tell whether low would have solved a task without rerunning it — and you pay full price for every token either way.
Related event: Max reasoning effort costs 42x more, often unnecessary(4 posts)→
More from coding & agent
- GitHub open-sources Replica, an AI prototyping tool built on your design system — round · 2026-10-29
- Agentic debugging framed as a constraint satisfaction problem in viral essay — goyal__pramod · 2026-10-29
- Pydantic webinar: trace coding agents in Logfire to compare cost and quality — samuelcolvin · 2026-10-29
- ChatGPT's dot can now text you via iMessage, with per-channel routing tips — nickbaumann_ · 2026-10-29
- Prevention beats repair: proactive AI saves ~$225K/year vs $120K for fix-it agents — louisdorard · 2026-10-29
- Product engineering in 2026: let Claude Code read the meeting transcript and build — curious_vii · 2026-10-29