Benchmarks · coding
Senior SWE-Bench
Uses the first-party tasteful solve rate (pass@1) at each model's published effort. Score: tasteful solve rate (pass@1). Collected by Jake Boggs' benchmark dashboard.
23 models · 23 results · data as of 2026-11-06 · senior-swe-bench.snorkel.ai ↗
Leaderboard
Opus 5.5 max
All results
23 rows| # | Model | Effort | Senior SWE-Bench |
|---|---|---|---|
| 1 | Opus 5.5 | max | 39.3% |
| 2 | GPT-6 Astra | max | 36.8% |
| 3 | Fable 5.1 | max | 34.7% |
| 4 | Fable 5 | max | 34.7% |
| 5 | Opus 5 | max | 34.7% |
| 6 | GPT-5.6 Sol | max | 34.7% |
| 7 | Opus 4.8 | max | 30.5% |
| 8 | GPT-5.5 | xhigh | 29.5% |
| 9 | Sonnet 5 | max | 28.0% |
| 10 | Gemini 4 Argon | max | 27.7% |
| 11 | GPT-5.6 Terra | max | 27.4% |
| 12 | Grok 4.7 | max | 27.4% |
| 13 | Grok 4.6 | max | 26.3% |
| 14 | Grok 4.5 | max | 23.2% |
| 15 | DeepSeek V4.1 Flash | max | 22.7% |
| 16 | MiniMax M3 | max | 21.1% |
| 17 | Kimi K3 | max | 20.2% |
| 18 | Opus 4.7 | max | 18.1% |
| 19 | GLM 5.2 | max | 17.9% |
| 20 | GPT-5.4 | xhigh | 16.8% |
| 21 | Gemini 3.7 Flash | max | 14.7% |
| 22 | Gemini 3.5 Flash | xhigh | 6.3% |
| 23 | Gemini 3.1 Pro | high | 2.1% |