Opus 5.5 benchmark scores top Fable 5.1 and GPT-6 Astra

Ananth @Ananth7eOriginally posted 2026.09.23

Open original on X

i just hope its not benchmaxxed like opus 5 was. it beats Fable 5.1 and GPT-6 Astra on most of the benchmarks. > Terminal-Bench 4.0 → 66.4% > FrontierCode → 54.4% > CursorBench → 57.8% if this model is as good in real world use as the benchmarks suggest, then anthropic kinda cooked.

This post quotes another post: x.com/synthwavedd/status/2102434467868799100