Benchmark
SWE-bench Verified
Coding benchmarks compare models on programming tasks. Scores are percentages. Higher is better.
- Models measured
- 130
- Scores recorded
- 205
- Sources
- 5
- IQ weight
- 2.70
Score distribution
- 70%
- Mean
- 73.5%
- Median
- 13.7%
- Std dev
- 51.6%
- 10th percentile
- 80.8%
- 90th percentile
- 8.7%
- Min
This benchmark has a weight of 2.70 for chat models, measured over 129 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 51 | MiMo V2.5 Pro· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 78.9% |
| 52 | MiMo V2.5 ProXiaomi MiMo-V2.5-Pro official launch table · published coding-agent setting· Reported by Xiaomi MiMo· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 78.9% |
| 53 | Qwen3.6 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 78.8% |
| 54 | Qwen3.6 PlusQwen official release table · thinking mode unless benchmark requires otherwise · internal bash + file-edit scaffold · 200K · temp 1.0 · top_p .95· Reported by Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 78.8% |
| 55 | DeepSeek V4 FlashDeepSeek V4 official model card · Think High· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 78.6% |
| 56 | Gemini 3 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 78% |
| 57 | Gemini 3 Flashsingle attempt · internal bash/file/submit scaffold · mean of 5 runs· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 78% |
| 58 | Hy3 (Deep Infra)· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 78% |
| 59 | Hy3 (Deep Infra)Hy3 official Benchmark Appendix · highest reasoning tier unless noted· Reported by Tencent Hunyuan· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 78% |
| 60 | MiMo V2 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 78% |
| 61 | MiMo V2 ProXiaomi MiMo-V2-Pro release table · harness unspecified· Reported by Xiaomi MiMo· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 78% |
| 62 | GLM 5· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 4 sources
| 77.8% |
| 63 | GLM 5GLM-5 official release table· Reported by Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 77.8% |
| 64 | Qwen3.7 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 77.7% |
| 65 | Qwen3.7 PlusQwen3.7-Plus release · internal bash + file-edit agent scaffold · temp 1.0 · top_p 0.95 · 200K· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 77.7% |
| 66 | Inkling FP4· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 3 sources
| 77.6% |
| 67 | Mistral Medium 3.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 77.6% |
| 68 | Mistral Medium 3.5 128b· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 77.6% |
| 69 | Muse Spark· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 77.4% |
| 70 | Claude Sonnet 4.5Opus 4.6 System Card capability summary · adaptive thinking · max effort · default sampling · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 77.2% |
| 71 | Claude Sonnet 4.5standard evaluation· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 77.2% |
| 72 | Qwen3.6 27B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 77.2% |
| 73 | Qwen3.6 27BQwen official release table · thinking mode unless benchmark requires otherwise · internal bash + file-edit scaffold · 200K· Reported by Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 77.2% |
| 74 | Kimi K2.5setting not named by the source· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76.8% |
| 75 | Kimi K2.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 4 sources
| 76.8% |
| 76 | Doubao Seed 2.0 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 76.5% |
| 77 | Qwen3.5 397B A17B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 3 sources
| 76.4% |
| 78 | Qwen3.5 397B A17BQwen3.5 release/model card · published setting· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76.4% |
| 79 | GPT 5.1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 76.3% |
| 80 | GPT-5.1-Instant· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 76.3% |
| 81 | Gemini 3 Pro Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 76.2% |
| 82 | Gemini 3 Pro PreviewGemini 3 Pro Model Card · results as of November 2025 · single attempt· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76.2% |
| 83 | Gemini 3 Pro PreviewOpus 4.6 System Card capability summary · adaptive thinking · max effort · default sampling · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76.2% |
| 84 | MiniMax M2.5official prose · OpenCode · default prompt· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76.1% |
| 85 | Muse Glimmer 30B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 76% |
| 86 | Muse Glimmer 30Bhigh reasoning· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76% |
| 87 | GPT 5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 74.9% |
| 88 | MiMo V2 Omni· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 74.8% |
| 89 | MiMo V2 OmniXiaomi MiMo-V2-Omni release table · code-and-text agent table · harness unspecified· Reported by Xiaomi MiMo· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 74.8% |
| 90 | Laguna M.1· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 74.6% |
| 91 | Claude Opus 4.1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 74.5% |
| 92 | GPT-5 Codex· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 74.5% |
| 93 | Step 3.5 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 3 sources
| 74.4% |
| 94 | Step 3.5 FlashStep 3.5 Flash official headline table · vanilla· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 74.4% |
| 95 | MiniMax M2.1· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 74% |
| 96 | Ring 2.6 1T· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 74% |
| 97 | GLM 4.7· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 73.8% |
| 98 | DeepSeek V4 FlashDeepSeek V4 official model card · Non-Think· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 73.7% |
| 99 | GPT-5.1 Codex· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 73.7% |
| 100 | Mai Thinking 1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 73.5% |
Showing 51–100 of 205 results.
Data credits: Data by LLM Stats · Hugging Face official benchmark leaderboards; licence varies by benchmark · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Steel.dev Agent Leaderboards, MIT (licence) · OpenEvals/leaderboard-data, MIT (licence)