← Benchmarks
SWE-Bench Pro
SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.
id swe-bench-pro · max 1 · 59 models reported
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 5.5 Anthropic | 0.90 |
| 2 | Claude Fable 5 Anthropic | 0.80 |
| 3 | Claude Mythos Preview Anthropic | 0.78 |