Category ranking · 6 models · checked 2026-10-13
Coding
Ranked by SWE-bench Verified, % resolved. Source: Model cards (vendor-reported, each with its own agent scaffold).
SWE-bench Verified, % resolved
2 with a published score, 4 not yet measured.
# | Model | Parameters | Score | Source | Licence |
|---|---|---|---|---|---|
| 1 | Devstral Small 1.1 Mistral AI | 23.6 B | 53.6%Sourced | Source SWE-Bench Verified with the OpenHands scaffold, model card. | Apache-2.0 |
| 2 | OpenHands LM 32B All Hands AI | 32.8 B | 37.2%Sourced | Source Model card, iterative evaluation protocol. | MIT |
| — | DeepSeek Coder V2 Lite DeepSeek | 15.7 B | Not yet measured | — | DeepSeek License |
| — | Qwen2.5 Coder 32B Alibaba Qwen | 32.8 B | Not yet measured | — | Apache-2.0 |
| — | Qwen2.5 Coder 7B Alibaba Qwen | 7.6 B | Not yet measured | — | Apache-2.0 |
| — | Qwen3 Coder 30B-A3B Alibaba Qwen | 30.5 B | Not yet measured | The card reports results as an image only; no text value captured. | Apache-2.0 |
How to read this: one public benchmark per category, so you can compare like with like. We copy the published number and link it; we do not re-run it, average different benchmarks or fill gaps. Vendor-reported numbers use each vendor's own settings, so treat small gaps as ties. “Not yet measured” means we found no comparable public score, not that the model is weak.
Every model was checked against the Hugging Face API on 2026-10-13: it exists, is public, and has the parameter count and licence shown. Other categories: Math and reasoning · Vision-language · OCR and documents · Speech to text · Text to speech · Image generation · Embeddings and rerankers · Small fast 'System 1' · Biomedical · Tool use and agents · Multilingual and translation · Time series · Robotics (VLA).
Machine-readable: /api/v1/categories/coding.json