SRE Bench
Reverse engineering real-world, contamination-free binaries. Run by Vals AI.
| Category | Agents |
|---|---|
| Metric | Accuracy (%) |
| Source | Vals AI (evaluator) |
| License | None stated |
| Benchmark page | www.vals.ai |
| Status | Scored |
| Weight | 0.9 within Agents |
| Fit | Discrimination 3.48, noise 0.72 |
Results
| # | Model | Provider | Result | Date |
|---|---|---|---|---|
| 1 | GPT-6 Astra Max | OpenAI | 56.9% | 2026-11-05 |
| 2 | GPT-6.1 Sol Max | OpenAI | 50.8% | 2026-11-05 |
| 3 | Gemini 4 Argon High | 44.3% | 2026-11-05 | |
| 4 | Claude Opus 5.5 Max | Anthropic | 33.6% | 2026-11-05 |
| 5 | GPT-5.6 Sol Max | OpenAI | 31.7% | 2026-11-05 |
| 6 | GPT-6 Sol Max | OpenAI | 30.5% | 2026-11-05 |
| 7 | Claude Sonnet 5.5 Max | Anthropic | 30.2% | 2026-11-05 |
| 8 | Kimi K3.1 Pro Max | Moonshot AI | 27.1% | 2026-11-05 |
| 9 | Claude Fable 5.1 Max | Anthropic | 22.9% | 2026-11-05 |
| 10 | Claude Opus 5 Max | Anthropic | 12.2% | 2026-11-05 |
| 11 | Muse Wonder Full Max | Meta | 11.8% | 2026-11-05 |
| 12 | Gemini 3.8 Flash High | 11.1% | 2026-11-05 | |
| 13 | GPT-5.6 Terra Max | OpenAI | 9.5% | 2026-11-05 |
| 14 | GLM-5.3 Max | Zhipu AI | 9.2% | 2026-11-05 |
| 15 | Claude Haiku 5.5 Max | Anthropic | 5.7% | 2026-11-05 |
| 16 | Gemini 3.7 Flash High | 4.6% | 2026-11-05 | |
| 17 | MiMo-V2.6-Flash | Xiaomi | 4.2% | 2026-11-05 |
| 18 | Grok 4.7 Extra High | xAI | 3.8% | 2026-11-05 |
| 19 | GPT-5.5 Extra High | OpenAI | 3.8% | 2026-11-05 |
| 20 | Muse Spark 1.3 Max | Meta | 3.8% | 2026-11-05 |
| 21 | MiMo-V2.6-Pro | Xiaomi | 3.1% | 2026-11-05 |
| 22 | GPT-6 Luna Max | OpenAI | 2.7% | 2026-11-05 |
| 23 | Hy4 preview | Tencent | 2.3% | 2026-11-05 |
| 24 | Inkling 2 Small | Thinking Machines | 1.9% | 2026-11-05 |
| 25 | DeepSeek V4 Pro 0813 Max | DeepSeek | 1.9% | 2026-11-05 |
| 26 | GPT-5.6 Luna Max | OpenAI | 1.5% | 2026-11-05 |
| 27 | DeepSeek V4.1 Flash Max | DeepSeek | 0.8% | 2026-11-05 |
| 28 | GLM-5.3-Flash Max | Zhipu AI | 0.8% | 2026-11-05 |
| 29 | Grok 4.5 High | xAI | 0.8% | 2026-11-05 |
| 30 | Step 5 Preview | StepFun | 0.4% | 2026-11-05 |
| 31 | Claude Sonnet 5 Max | Anthropic | 0.4% | 2026-11-05 |
| 32 | Inkling | Thinking Machines | 0.0% | 2026-11-05 |
| 33 | Gemini 3.1 Pro Preview High | 0.0% | 2026-11-05 | |
| 34 | GLM-5.2 Max | Zhipu AI | 0.0% | 2026-11-05 |
| 35 | Muse Spark 1.2 Extra High | Meta | 0.0% | 2026-11-05 |