Code Generation on LiveCodeBench v6 (Accuracy)
87.4AccuracyGemini 3 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 3 ProParams=N/A2026.06 | 87.4 | |
| Qwen3.6 PlusParams=N/A2026.06 | 87.1 | |
| GLM-5Params=744B2026.06 | 85.5 | |
| Kimi K2.5Params=1T2026.06 | 85 | |
| Claude Opus 4.5Params=N/A2026.06 | 84.8 | |
| GPT-5 (high)Params=N/A2026.06 | 84.5 | |
| Qwen3.5-397B-A17BParams=397B2026.06 | 83.6 | |
| GPT-OSS (high)Params=120B2026.06 | 81.9 | |
| DeepSeek V3.2Params=671B2026.06 | 80.8 | |
| MiMo v2 FlashParams=309B2026.06 | 80.6 | |
| VibeThinker-3BParams=3B2026.06 | 80.2 | |
| OpenAI o3 (high)Params=N/A2026.06 | 75.8 | |
| Gemini 2.5 ProParams=N/A2026.06 | 72.5 | |
| DeepSeek R1 0528Params=671B2026.06 | 68.7 | |
| Gemini 2.5 FlashParams=N/A2026.06 | 61.2 |