EDIT-Bench Complete Easy
Code · 2025-11-06
The EDIT-Bench Complete easy slice reports pass@1 on the 295 tasks classified as easy by the benchmark authors.
Top models (higher is better)
| Model | Score |
|---|---|
| Sonnet 4 | 93.6 |
| Sonnet 4.5 | 93.6 |
| Claude 3.7 Sonnet | 90.8 |
| Claude 3.5 Sonnet (Oct 2024) | 86.8 |
| o3-mini | 86.8 |
| GLM 4.6 | 85.1 |
| O4 Mini | 85.1 |
| Kimi K2 Instruct 0905 | 84.4 |
| Grok 4 Fast | 83.0 |
| GPT-5 Mini | 82.7 |
| Qwen3-Coder-480B-A35B-Instruct | 81.7 |
| GPT-4o (2024-08-06) | 81.4 |
| DeepSeek-V3.1 | 81.0 |
| Grok Code Fast 1 | 80.7 |
| Llama 3.3 70B Instruct | 79.3 |
| Qwen3 14B | 79.0 |
| GPT-5 | 77.3 |
| Llama 4 Maverick Instruct | 77.3 |
| GPT-4o Mini | 76.6 |
| Llama 3.1 405B Instruct | 76.6 |
| Qwen3 Coder Flash (rolling alias) | 76.3 |
| gpt-oss-20b | 75.9 |
| Gemini 2.5 Pro | 75.6 |
| Qwen2.5 72B Instruct | 73.2 |
| Mistral Small 3.2 24B Instruct 2506 | 72.5 |
| Qwen3-30B-A3B | 72.5 |
| GPT-5 Nano | 71.9 |
| Codestral 25.08 | 70.8 |
| DeepSeek-R1-0528 | 69.5 |
| Llama 4 Scout Instruct | 69.2 |
| Devstral Medium 1.0 | 67.5 |
| gpt-oss-120b | 64.4 |
| Qwen2.5 Coder 32B Instruct | 63.7 |
| Devstral Small 1.1 | 60.3 |
| Gemma 3 27B IT | 59.0 |
| Llama 3.1 8B Instruct | 54.9 |
| Kimi-Dev-72B | 53.2 |
| Gemma 3 12B IT | 49.5 |
| Gemma 3n E4B Instruct | 47.8 |
| GLM-4.5 | 45.8 |