Benchmarks / OpenHands Index
Reported by OpenHands Index
OpenHands Index
Building a library from its specification until its test suite passes.
- Results dated
- 30 Jun 2026
- Models
- 34
- Unit
- % resolved
- Licence
- Apache License 2.0
| # | Model | OpenHands Index: greenfield (Commit0) % resolved, higher is better |
|---|---|---|
| 1 | Claude Fable 5Anthropic · claude-fable-5 |
62.5%
|
| 1 | Claude Opus 4.8Anthropic · claude-opus-4.8 |
62.5%
|
| 3 | Claude Opus 4.6Anthropic · claude-opus-4.6 |
56.2%
|
| 3 | Claude Opus 4.7Anthropic · claude-opus-4.7 |
56.2%
|
| 3 | GPT-5.4OpenAI · gpt-5.4 |
56.2%
|
| 6 | Claude Sonnet 4.6Anthropic · claude-sonnet-4.6 |
50.0%
|
| 6 | GPT-5.2OpenAI · gpt-5.2 |
50.0%
|
| 8 | GPT-5.2-CodexOpenAI · gpt-5.2-codex |
43.8%
|
| 8 | GPT-5.5OpenAI · gpt-5.5 |
43.8%
|
| 10 | Claude Opus 4.5Anthropic · claude-opus-4.5 |
37.5%
|
| 10 | Gemini 3.5 FlashGoogle · gemini-3.5-flash |
37.5%
|
| 10 | GLM 5.1Z.ai · glm-5.1 |
37.5%
|
| 13 | GLM 5Z.ai · glm-5 |
31.2%
|
| 14 | Qwen3.6 PlusAlibaba · qwen3.6-plus |
25.0%
|
| 14 | Qwen3 Coder NextAlibaba · qwen3-coder-next |
25.0%
|
| 14 | DeepSeek V3.2DeepSeek · deepseek-v3.2 |
25.0%
|
| 14 | Gemini 3.1 Pro PreviewGoogle · gemini-3.1-pro-preview |
25.0%
|
| 14 | gemini-3-proGoogle |
25.0%
|
| 14 | MiniMax M3MiniMax · minimax-m3 |
25.0%
|
| 14 | Kimi K2.6Moonshot AI · kimi-k2.6 |
25.0%
|
| 21 | gemini-3-flashGoogle |
18.8%
|
| 21 | MiniMax M2.1MiniMax · minimax-m2.1 |
18.8%
|
| 21 | MiniMax M2.7MiniMax · minimax-m2.7 |
18.8%
|
| 21 | Kimi K2.5Moonshot AI · kimi-k2.5 |
18.8%
|
| 25 | Qwen3.5-FlashAlibaba · qwen3.5-flash-02-23 |
12.5%
|
| 25 | Claude Sonnet 4.5Anthropic · claude-sonnet-4.5 |
12.5%
|
| 25 | Trinity Large ThinkingArcee Ai · trinity-large-thinking |
12.5%
|
| 25 | DeepSeek V4 Pro 0423DeepSeek · deepseek-v4-pro |
12.5%
|
| 25 | MiniMax M2.5MiniMax · minimax-m2.5 |
12.5%
|
| 25 | Kimi K2 ThinkingMoonshot AI · kimi-k2-thinking |
12.5%
|
| 25 | Nemotron 3 SuperNVIDIA · nemotron-3-super-120b-a12b |
12.5%
|
| 25 | GLM 4.7Z.ai · glm-4.7 |
12.5%
|
| 33 | Nemotron 3 Nano 30B A3BNVIDIA · nemotron-3-nano-30b-a3b |
6.2%
|
| 34 | Qwen3 Coder 480B A35BAlibaba · qwen3-coder |
0.0%
|
Results as published by OpenHands Index; we do not re-run them.
What it measures
Building a library from its specification until its test suite passes.
What it does not measure
Not the model alone: the score is for the model inside the OpenHands agent, at the SDK version shown.
Source
OpenHands Index by the OpenHands contributors, Apache License 2.0.