Internal evals: GPT-6.1-sol hits 3x the efficiency of Opus 5.5 in surprise point release

charliermarsh · x · 2026-09-30

Founder blader shared results from internal evals at cfo.ai: while Opus 5.5 previously measured 2x as efficient as GPT-6-sol, the new 6.1-sol point release reversed the gap — reaching roughly 3x the efficiency of Opus 5.5 and 2.5x that of Sonnet 5.5. He calls it an "insane point release". Note these are third-party internal benchmarks, not official figures.

Related event: Internal Evals Claim GPT-6.1 Sol Beats Claude Opus 5.5 on Cost and Speed(5 posts)→

Original post →

More from Models

Models channel →