CyberBench v1.1
Crafting proof-of-concept inputs that trigger OSS-Fuzz vulnerabilities and writing source patches that fix them. Run by Vals AI.
| Category | Agents |
|---|---|
| Metric | Accuracy (%) |
| Source | Vals AI (evaluator) |
| License | None stated |
| Benchmark page | www.vals.ai |
| Status | Scored |
| Weight | 0.8 within Agents |
| Fit | Discrimination 0.08, noise 0.53 |
Results
| # | Model | Provider | Result | Date |
|---|---|---|---|---|
| 1 | Muse Wonder Max | Meta | 80.4% | 2026-11-05 |
| 2 | GPT-6 Sol Max | OpenAI | 78.0% | 2026-11-05 |
| 3 | Gemini 4 Argon High | 77.9% | 2026-11-05 | |
| 4 | GPT-5.6 Sol Max | OpenAI | 76.3% | 2026-11-05 |
| 5 | GPT-6 Luna Max | OpenAI | 76.3% | 2026-11-05 |
| 6 | Claude Haiku 5.5 Max | Anthropic | 75.5% | 2026-11-05 |
| 7 | MiMo-V2.6-Flash | Xiaomi | 75.4% | 2026-11-05 |
| 8 | Kimi K3 Max | Moonshot AI | 75.2% | 2026-11-05 |
| 9 | DeepSeek V4.1 Flash High | DeepSeek | 73.7% | 2026-11-05 |
| 10 | GPT-5.6 Luna Max | OpenAI | 73.6% | 2026-11-05 |
| 11 | Muse Wonder Full Max | Meta | 73.6% | 2026-11-05 |
| 12 | MiMo-V2.6-Pro | Xiaomi | 72.9% | 2026-11-05 |
| 13 | DeepSeek V4 Pro 0813 Max | DeepSeek | 72.9% | 2026-11-05 |
| 14 | Inkling 2 Small | Thinking Machines | 72.7% | 2026-11-05 |
| 15 | Muse Spark 1.3 Max | Meta | 72.7% | 2026-11-05 |
| 16 | GLM-5.3 Max | Zhipu AI | 72.1% | 2026-11-05 |
| 17 | GPT-5.6 Terra Max | OpenAI | 72.1% | 2026-11-05 |
| 18 | DeepSeek V4 Flash 0731 High | DeepSeek | 72.1% | 2026-11-05 |
| 19 | Grok 4.5 High | xAI | 71.0% | 2026-11-05 |
| 20 | Claude Fable 5.1 Max | Anthropic | 70.4% | 2026-11-05 |
| 21 | Ember-1 | Fireworks | 69.5% | 2026-11-05 |
| 22 | Muse Spark 1.2 Extra High | Meta | 69.5% | 2026-11-05 |
| 23 | Grok 4.7 Extra High | xAI | 69.5% | 2026-11-05 |
| 24 | Muse Spark 1.3 Extra High | Meta | 69.4% | 2026-11-05 |
| 25 | GLM-5.3-Flash Max | Zhipu AI | 67.9% | 2026-11-05 |
| 26 | Qwen3.8 27B Extra High | Alibaba | 67.8% | 2026-11-05 |
| 27 | Grok 4.6 High | xAI | 67.7% | 2026-11-05 |
| 28 | Inkling | Thinking Machines | 66.8% | 2026-11-05 |
| 29 | Hy4 preview | Tencent | 65.4% | 2026-11-05 |
| 30 | Claude Opus 5 Max | Anthropic | 65.4% | 2026-11-05 |
| 31 | MiniMax-M3 | MiniMax | 65.0% | 2026-11-05 |
| 32 | Inkling Small | Thinking Machines | 64.1% | 2026-11-05 |
| 33 | Claude Sonnet 5 Max | Anthropic | 61.9% | 2026-11-05 |
| 34 | Claude Sonnet 5.5 Max | Anthropic | 59.6% | 2026-11-05 |
| 35 | Gemini 3.5 Flash Lite High | 59.1% | 2026-11-05 | |
| 36 | Claude Opus 5.5 Max | Anthropic | 55.4% | 2026-11-05 |
| 37 | Mercury 2.5 High | Inception | 45.5% | 2026-11-05 |
| 38 | Gemini 3.6 Flash High | 45.4% | 2026-11-05 | |
| 39 | Gemini 3.8 Flash High | 43.8% | 2026-11-05 | |
| 40 | Gemini 3.7 Flash High | 43.8% | 2026-11-05 | |
| 41 | GPT-6 Astra Max | OpenAI | 41.1% | 2026-11-05 |
| 42 | GPT-6.1 Sol Max | OpenAI | 39.3% | 2026-11-05 |
| 43 | Qwen3.8 Max | Alibaba | 28.6% | 2026-11-05 |