Claude Fable 5.1 Leads FrontierSWE v2 Benchmark by 24 Percentage Points in 20 Hour Test

The latest autonomous coding evaluation from Proximal shows a wide performance gap between top AI models. Claude Fable 5.1 outperformed all other frontier mode…