Competitive Programming on LiveCodeBench Pro 2025Q2
71.9Pass@1Gemini-3-pro
Evaluation results
| Method | Links | |
|---|---|---|
| Gemini-3-proBackbone=Gemini-3-pro2026.05 | 71.9 | |
| o4-mini-highBackbone=o4-mini-high2026.05 | 59.3 | |
| DeepSeek-V3.2-ReasonerBackbone=DeepSeek-V3.2-Reasoner2026.05 | 58.1 | |
| CP-AgentBackbone=DeepSeek-V3.2-Chat2026.05 | 48.5 | |
| CP-AgentBackbone=Qwen3-235B-A22B-Instruct2026.05 | 46.7 | |
| CodeSimBackbone=Qwen3-235B-A22B-Instruct2026.05 | 45.5 | |
| CodeSimBackbone=DeepSeek-V3.2-Chat2026.05 | 40.1 | |
| DeepSeek-R1Backbone=DeepSeek-R12026.05 | 39.5 | |
| ReflexionBackbone=DeepSeek-V3.2-Chat2026.05 | 37.1 | |
| ReflexionBackbone=Qwen3-235B-A22B-Instruct2026.05 | 35.3 | |
| Qwen3-235B-A22B-InstructBackbone=Qwen3-235B-A22B-Instruct2026.05 | 26.9 | |
| AlphaCodiumBackbone=Qwen3-235B-A22B-Instruct2026.05 | 25.8 | |
| DeepSeek-V3.2-ChatBackbone=DeepSeek-V3.2-Chat2026.05 | 25.8 | |
| AlphaCodiumBackbone=DeepSeek-V3.2-Chat2026.05 | 24.5 | |
| CP-AgentBackbone=GPT-4o2026.05 | 10.2 | |
| ReflexionBackbone=GPT-4o2026.05 | 9.6 | |
| CodeSimBackbone=GPT-4o2026.05 | 8.4 | |
| GPT-4oBackbone=GPT-4o2026.05 | 5.4 | |
| AlphaCodiumBackbone=GPT-4o2026.05 | 5.4 |