Coding on LiveCodeBench v5
77.6AccuracyQwen3-235B-A22B-R-TAP
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen3-235B-A22B-R-TAPArchitecture=MoE, Parameter Count=235B2026.03 | 77.6 | |
| AM-Thinking-v1-R-TAPArchitecture=Dense, Parameter Count=32B2026.03 | 76.5 | |
| Gemini 2.5 Pro2026.02 | 75.9 | |
| Teacher Bestinference_mode=thinking mode2026.06 | 73.74 | |
| Qwen3-32B-R-TAPArchitecture=Dense, Parameter Count=32B2026.03 | 71 | |
| Qwen3-235B-A22BArchitecture=MoE, Parameter Count=235B2026.03 | 70.7 | |
| Gemini2.5-Pro2026.03 | 70.4 | |
| AM-Thinking-v1Architecture=Dense, Parameter Count=32B2026.03 | 70.3 | |
| MOPDinference_mode=thinking mode2026.06 | 69.53 | |
| Nemotron-Nano-v2# Param.=9B2026.02 | 68.26 | |
| Nemetron-Ultra-253BArchitecture=Dense, Parameter Count=256B2026.03 | 68.1 | |
| Falcon-H1R# Param.=7B2026.02 | 67.66 | |
| OpenAI-o3-mini (Medium)Configuration=Medium2026.03 | 66.3 | |
| Ministral-3-R# Param.=8B2026.02 | 65.87 | |
| Qwen3-8Binference_mode=thinking mode2026.06 | 65.78 | |
| Qwen3-32BArchitecture=Dense, Parameter Count=32B2026.03 | 65.7 | |
| Claude Sonnet 4Reasoning=Thinking2026.02 | 65.5 | |
| Base (SFT)inference_mode=thinking mode2026.06 | 65.45 | |
| DeepSeek-R1Architecture=MoE, Parameter Count=671B2026.03 | 64.3 | |
| OpenAI-o1 (2024-12-17)2026.03 | 63.9 | |
| Claude Opus 4Reasoning=Thinking2026.02 | 63.6 | |
| MiniCPM-SALA# Param.=9B2026.02 | 60.48 | |
| Interfaze-Beta2026.02 | 57.77 | |
| Qwen3# Param.=8B2026.02 | 56.89 | |
| MiniCPM-4.1# Param.=8B2026.02 | 56.89 | |
| GPT-5Reasoning=Minimal Reasoning2026.02 | 55.8 | |
| Gemini 2.5 Flash2026.02 | 49.5 | |
| GPT-4.12026.02 | 45.7 | |
| Claude Sonnet 42026.02 | 44.9 | |
| Qwen3-1.7BModel Size=1.7B, Architecture=Standard, Few-shot setting=0-shot, Max tokens=327682026.02 | 25.5 | |
| Qwen3-1.7B-ALLMEMModel Size=1.7B, Architecture=ALLMEM, Sliding window length=2048, Attention sinks=128, Few-shot setting=0-shot, Max tokens=327682026.02 | 25 | |
| Qwen3-0.6B-ALLMEMModel Size=0.6B, Architecture=ALLMEM, Sliding window length=2048, Attention sinks=128, Few-shot setting=0-shot, Max tokens=327682026.02 | 14.3 | |
| Qwen3-0.6BModel Size=0.6B, Architecture=Standard, Few-shot setting=0-shot, Max tokens=327682026.02 | 13.8 |