DeepSWE 1.1
25 models
Top 5 Models Performance
| Google: gemini-4-argon | ######################################## | 77.9 |
| DeepSeek: deepseek-v4.1-flash-1017 | ######################################## | 77 |
| Anthropic: claude-opus-5.5 | ###################################### | 74.2 |
| DeepSeek: deepseek-v4.1-flash | ###################################### | 74.2 |
| OpenAI: gpt-6-astra | ###################################### | 74.1 |
| Rank | Model | Score | |
|---|---|---|---|
| 🇺🇸 | 1 | Google: gemini-4-argon | 77.9 |
| 🇨🇳 | 2 | DeepSeek: deepseek-v4.1-flash-1017深度求索 | 77 |
| 🇺🇸 | 3 | Anthropic: claude-opus-5.5 | 74.2 |
| 🇨🇳 | 4 | DeepSeek: deepseek-v4.1-flash深度求索 | 74.2 |
| 🇺🇸 | 5 | OpenAI: gpt-6-astra | 74.1 |
| 🇺🇸 | 6 | OpenAI: gpt-5.6-sol | 73 |
| 🇺🇸 | 7 | Anthropic: Claude Fable 5 | 70 |
| 🇺🇸 | 8 | OpenAI: gpt-5.6-terra | 70 |
| 🇺🇸 | 9 | Anthropic: claude-opus-5 | 68.8 |
| 🇺🇸 | 10 | OpenAI: GPT-5.5 | 67 |
| 🇺🇸 | 11 | OpenAI: gpt-5.6-luna | 67 |
| 🇨🇳 | 12 | Z.ai: glm-5.3北京智谱华章科技有限公司 | 66.9 |
| 🇺🇸 | 13 | SpaceXAI: grok-4.6 | 65.9 |
| 🇺🇸 | 14 | Meta: muse-spark-open-816b-a42b | 60 |
| 🇺🇸 | 15 | Meta: muse-spark-1.2 | 59.3 |
| 🇺🇸 | 16 | Anthropic: Claude Opus 4.8 | 59 |
| 🇨🇳 | 17 | Qwen: qwen3.8-flash-next通义千问 | 58.7 |
| 🇺🇸 | 18 | Anthropic: claude-sonnet-5 | 54 |
| 🇺🇸 | 19 | Meta: muse-spark-1.1 | 53.3 |
| 🇺🇸 | 20 | OpenAI: GPT-5.4 Pro | 52 |
| ❓ | 21 | recursiveai: ripple-65b-preview | 44.3 |
| 🇨🇳 | 22 | Z.ai: GLM 5.2北京智谱华章科技有限公司 | 44 |
| 🇨🇳 | 23 | Qwen: qwen3.8-27b通义千问 | 42.2 |
| 🇺🇸 | 24 | Google: Gemini 3.5 Flash | 37 |
| 🇺🇸 | 25 | Meta: Muse Spark | 10 |