Reasoning on LiveCodeBench v6
90.7ScoreGemini-3.0 Pro
Evaluation results
| Method | Links | |
|---|---|---|
| Gemini-3.0 ProVariant=Pro2026.01 | 90.7 | |
| Audex 30B-A3BContext Length=1M, Tool-integrated reasoning=true2026.07 | 86.2 | |
| Audex 30B-A3BContext Length=1M, Tool-integrated reasoning=false2026.07 | 85.3 | |
| MiMo-V2-FlashVariant=Flash2026.01 | 85.1 | |
| GPT-5 HighVariant=High2026.01 | 84.5 | |
| DeepSeek-V3.2 ThinkingThinking Mode=true2026.01 | 83.3 | |
| Kimi-K2 ThinkingThinking Mode=true2026.01 | 83.1 | |
| EXAONE 4.5 33BArchitecture=Dense, # Total Params=33B, # Activated Params=32B, Reasoning Mode=REASONING2026.04 | 81.4 | |
| K-EXAONE-236B-A23BArchitecture=MoE, # Total Params=236B, # Activated Params=23B, Reasoning Mode=REASONING2026.04 | 80.7 | |
| Qwen3.5-27BArchitecture=Dense, # Total Params=27B, # Activated Params=27B, Reasoning Mode=REASONING2026.04 | 80.7 | |
| GPT-5 miniReasoning Mode=REASONING: HIGH2026.04 | 78.1 | |
| Nemotron-3-Super 120B-A12BEvaluation Setting=non-reasoning2026.06 | 74.67 | |
| Kimi-K2.5Mode=Instant2026.06 | 73.4 | |
| Audex 2BContext Length=128K, Tool-integrated reasoning=true2026.07 | 72.7 | |
| GPT-5.4Reasoning Mode=non-reasoning2026.06 | 70.76 | |
| Qwen3-VL-235B-A22BArchitecture=MoE, # Total Params=236B, # Activated Params=22B, Reasoning Mode=Thinking2026.04 | 70.1 | |
| Ling-2.6-1T2026.06 | 65.58 | |
| Audex 2BContext Length=128K, Tool-integrated reasoning=false2026.07 | 64.5 | |
| Claude Sonnet 4.5Variant=Sonnet 4.52026.01 | 64 | |
| Ling-2.6-flash2026.06 | 62.28 | |
| GPT-OSS-120BEvaluation Setting=low2026.06 | 61.51 | |
| Qwen3-Omni 30B-A3B ThinkingContext Length=64K2026.07 | 59.2 | |
| DeepSeek-V3.2Thinking Mode=nothink2026.06 | 57.71 | |
| Qwen3.5-Omni Flash 35B-A3BContext Length=256K2026.07 | 56.6 | |
| GPT-5.4-miniEvaluation Setting=non-reasoning2026.06 | 55.07 | |
| GLM-5Thinking Mode=non-thinking2026.06 | 52.09 | |
| Step-Audio R1.1 33BContext Length=64K2026.07 | 22.9 | |
| Voxtral Small-24B 2507Context Length=128K2026.07 | 22.9 |