Benchmark
SWE-Bench Pro
Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages. Higher is better.
- Models and systems measured
- 93
- Scores recorded
- 188
- Sources
- 4
- IQ weight
- 3.48
Score distribution
- 57.3%
- Mean
- 57.7%
- Median
- 14%
- Std dev
- 46.3%
- 10th percentile
- 69.2%
- 90th percentile
- 5.2%
- Min
This benchmark has a weight of 3.48 for chat models, measured over 92 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model or system | Score |
|---|---|---|
| 101 | GPT 5.4MiniMax M2.7 official benchmark overview · exact harness/version not stated· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 57.7% |
| 102 | GPT 5.4Public· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 57.7% |
| 103 | Qwen3.7 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 57.6% |
| 104 | Qwen3.7 PlusQwen3.7-Plus release · internal bash + file-edit agent scaffold · temp 1.0 · top_p 0.95 · 200K · corrected/refined tasks· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 57.6% |
| 105 | Doubao Seed 2.1 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 57.5% |
| 106 | Claude Opus 4.6MiniMax M2.7 official benchmark overview · exact harness/version not stated· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 57.3% |
| 107 | Claude Sonnet 4.6MiniMax M2.7 official benchmark overview · exact harness/version not stated· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 57.2% |
| 108 | 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
| 57.2% |
| 109 | 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. | 57.2% |
| 110 | Doubao Seed 2.1 Turbo· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 57% |
| 111 | GPT-5.3 Codex· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 56.8% |
| 112 | GPT-5.3 CodexPublic· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 56.8% |
| 113 | GPT-5.3 CodexPublic · xhigh· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 56.8% |
| 114 | Ling 3.0 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 56.6% |
| 115 | Qwen3.6 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 56.6% |
| 116 | Qwen3.6 PlusQwen official release table · thinking mode unless benchmark requires otherwise · refined public set · 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. | 56.6% |
| 117 | GPT-5.2 Codex· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 56.4% |
| 118 | Step 3.7 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 56.3% |
| 119 | Step 3.7 FlashStep 3.7 Flash official README/figure· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 56.3% |
| 120 | MiniMax M2.7MiniMax M2.7 official benchmark overview · exact harness/version not stated· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 56.2% |
| 121 | MiniMax M2.7· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 56.2% |
| 122 | MiMo V2.5· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 56.1% |
| 123 | MiMo V2.5Xiaomi MiMo-V2.5 official release/model card · 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. | 56.1% |
| 124 | DeepSeek V4 Flash 0731Claude Code · temp 1.0 · top_p 0.95 · 256K · refined tasks;Claude 对照为官方值· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 56% |
| 125 | Inkling Small· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 55.9% |
| 126 | Qwen3.7 PlusClaude Code · temp 1.0 · top_p 0.95 · 256K · refined tasks;Claude 对照为官方值· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.8% |
| 127 | GPT 5.2Public· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.6% |
| 128 | GPT 5.2Public · xhigh· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.6% |
| 129 | DeepSeek V4 Pro· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 55.4% |
| 130 | MiniMax M2.5· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 55.4% |
| 131 | MiniMax M2.5main capability figure· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.4% |
| 132 | MiniMax M2.5MiniMax M2.7 official benchmark overview · exact harness/version not stated· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.4% |
| 133 | Gemini 3.5 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 55.1% |
| 134 | Gemini 3.5 FlashGemini 3.5 Flash Model Card · May 2026 · Public · single attempt· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.1% |
| 135 | Gemini 3.5 FlashGemini 3.6 Flash Model Card · July 2026 · Public· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.1% |
| 136 | GPT-5.4 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 54.4% |
| 137 | GPT-5.4 minixhigh · Public· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.4% |
| 138 | Inkling FP4· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 54.3% |
| 139 | Gemini 3.1 Pro Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 54.2% |
| 140 | Gemini 3.1 Pro PreviewFable/Mythos 5 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 141 | Gemini 3.1 Pro PreviewGemini 3.5 Flash Model Card · May 2026 · Public · single attempt· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 142 | Gemini 3.1 Pro PreviewGemini 3.6 Flash Model Card · July 2026 · Public· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 143 | Gemini 3.1 Pro Previewhigh · Scale AI · mini-swe-agent for Muse Spark 1.1· Reported by Meta, Meta Superintelligence Labs· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 144 | Gemini 3.1 Pro PreviewMiniMax M2.7 official benchmark overview · exact harness/version not stated· Reported by MiniMax· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 145 | Gemini 3.1 Pro PreviewOpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 146 | Gemini 3.1 Pro PreviewOpus 4.8 System Card capability summary · adaptive thinking · max effort · 5 trials unless noted· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 147 | Gemini 3.1 Pro Previewpublic subset · single attempt· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 54.2% |
| 148 | Gemini 3.5 Flash Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 54.2% |
| 149 | Ring 2.6 1T· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 53.8% |
| 150 | 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
| 53.5% |
Showing 101–150 of 187 results.
Not ranked
The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.
| Model or system | Score |
|---|---|
| Muse Spark (version unspecified)Meta Model API · highest effort · version unspecified · Scale AI · mini-swe-agent for Muse Spark 1.1· Reported by Meta, Meta Superintelligence Labs· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55% |
Data credits: Hugging Face official benchmark leaderboards; licence varies by benchmark · Data by LLM Stats · 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. · OpenEvals/leaderboard-data, MIT (licence)