Fable 5.1 Beats GPT-5.6 on Benchmark at Lower Cost

haider1 · x · 2026-09-02

Haider shared new results from CursorBench 3.2. Anthropic's Fable 5.1 Medium scored 68.0% at $3.53 per task, outperforming GPT-5.6 Sol Max which scored 67.2% at $5.69 per task. This indicates Fable now beats Sol on performance while being significantly cheaper. The author noted the efficiency jump arrived faster than expected.

Related event: Fable 5.1 Tops Benchmarks at Lower Cost(3 posts)→

Original post →

More from Models

Models channel →