Code Generation
LiveBench · 28 models
| Rank | Model | Score |
|---|---|---|
| 1 | claude-opus-4-7-xhigh-effort | 85.915 |
| 2 | claude-sonnet-5-xhigh-effort | 83.099 |
| 3 | gpt-5.5-xhigh | 81.69 |
| 4 | claude-fable-5-xhigh-effort | 80.282 |
| 4 | claude-opus-4-6-thinking-auto-high-effort | 80.282 |
| 4 | claude-opus-4-8-xhigh-effort | 80.282 |
| 4 | claude-sonnet-4-6-thinking-auto-medium-effort | 80.282 |
| 4 | gemini-3.5-flash-high | 80.282 |
| 4 | gpt-5.2-codex | 80.282 |
| 4 | qwen3.6-plus | 80.282 |
| 11 | claude-opus-4-5-20251101-thinking-64k-high-effort | 78.873 |
| 11 | glm-5.2 | 78.873 |
| 11 | kimi-k2.6-thinking | 78.873 |
| 11 | qwen3.7-max | 78.873 |
| 15 | gpt-5.5-high | 77.465 |
| 16 | gpt-5.2-2025-12-11-high | 76.056 |
| 17 | gemini-3.1-pro-preview-high | 74.648 |
| 17 | gpt-5.4-xhigh | 74.648 |
| 17 | gpt-5.5 | 74.648 |
| 17 | grok-4.3 | 74.648 |
| 21 | gpt-5.4-nano-xhigh | 74.286 |
| 22 | deepseek-v4-flash | 73.239 |
| 23 | kimi-k2.7-code | 71.831 |
| 23 | qwen3.6-27b | 71.831 |
| 25 | deepseek-v4-pro | 70.423 |
| 26 | minimax-m3 | 69.014 |
| 27 | gpt-5.4-mini-xhigh | 67.692 |
| 28 | grok-build-0.1 | 63.38 |