Code Generation on LiveCodeBench (LCB) (Pass@1)
91.7Pass@1Gemini-3-Pro
Evaluation results
| Method | Links | |
|---|---|---|
| Gemini-3-Pro2026.06 | 91.7 | |
| Claude-Opus-4.52026.06 | 87.1 | |
| GPT-5.12026.06 | 87 | |
| GLM-4.72026.06 | 84.9 | |
| DeepSeek-V3.22026.06 | 83.3 | |
| Qwen-7BFT data=OT3, LoRA=-2026.03 | 66 | |
| R1-Distill-Qwen-7BFT data=R1*, LoRA=-2026.03 | 59 | |
| Qwen-7BFT data=MoT, LoRA=-2026.03 | 55 | |
| Qwen-7BFT data=OT3, LoRA=rk 1282026.03 | 54 | |
| Qwen3-Coder-480B-A35B-Instruct2026.06 | 53.9 | |
| Kimi-K2-Instruct-09052026.06 | 53.7 | |
| Qwen3-235B-A22B-Instruct-25072026.06 | 51.8 | |
| Qwen-7BFT data=OT3 + MoT, LoRA=rk 1282026.03 | 51 | |
| Qwen-3BFT data=OT3, LoRA=-2026.03 | 50 | |
| Qwen-7BFT data=OT3, LoRA=rk 642026.03 | 48 | |
| Kimi-Dev-72B2026.06 | 40 | |
| Qwen-3BFT data=MoT, LoRA=-2026.03 | 38 | |
| Qwen-7BFT data=-, LoRA=-2026.03 | 36 | |
| Qwen3-14BBackbone=Qwen3-14B, Inference Template=chat inference template2025.05 | 35.8 | |
| INTUITORBackbone=Qwen3-14B, Inference Template=chat inference template2025.05 | 35.6 | |
| LoopCoder-v2R (number of refinement loops)=22026.06 | 35.4 | |
| INTUITORBackbone=Qwen2.5-14B, Inference Template=chat inference template2025.05 | 30 | |
| GRPOBackbone=Qwen2.5-14B, Inference Template=chat inference template2025.05 | 29.6 | |
| LoopCoder-v2R (number of refinement loops)=32026.06 | 28.6 | |
| Qwen2.5-Coder-32B-Instruct2026.06 | 27.4 | |
| LoopCoder-v2R (number of refinement loops)=12026.06 | 27.4 | |
| Qwen2.5-Coder-14B-Instruct2026.06 | 24.6 | |
| LoopCoder-v2R (number of refinement loops)=42026.06 | 24.5 | |
| Qwen-3BFT data=-, LoRA=-2026.03 | 23 | |
| Qwen-3BFT data=OT3, LoRA=rk 1282026.03 | 23 | |
| Seed-Coder-8B-Instruct2026.06 | 22.3 | |
| Qwen2.5-14BBackbone=Qwen2.5-14B, Inference Template=chat inference template2025.05 | 22 | |
| GRPOBackbone=Qwen2.5-7B, Inference Template=chat inference template2025.05 | 20 | |
| DeepSeek-Coder-V2-Lite-Instruct2026.06 | 19.4 | |
| INTUITORBackbone=Qwen2.5-7B, Inference Template=chat inference template2025.05 | 19 | |
| Qwen2.5-Coder-7B-Instruct2026.06 | 18.9 | |
| Qwen2.5-7BBackbone=Qwen2.5-7B, Inference Template=chat inference template2025.05 | 2.6 |