Models scored
49
evaluated
Modality
text
Category
agents
+3 more
Published
—
Search
| # | Model | Lab | Score |
|---|---|---|---|
| 01 | GPT-5.5 | OpenAI | 83 |
| 02 | Claude Mythos Preview | Anthropic | 82 |
| 03 | Claude Sonnet 5 | Anthropic | 80 |
| 04 | GPT-5.3 Codex | OpenAI | 77 |
| 05 | Gemini 3.5 Flash | 76 | |
| 06 | GPT-5.4 | OpenAI | 75 |
| 07 | Claude Opus 4.8 | Anthropic | 75 |
| 08 | Qwen3.7-Plus | Alibaba Cloud / Qwen Team | 70 |
| 09 | Qwen3.7 Max | Alibaba Cloud / Qwen Team | 70 |
| 10 | Claude Opus 4.7 | Anthropic | 69 |
| 11 | GLM-5.1 | Zhipu AI | 69 |
| 12 | Gemini 3.1 Pro | 69 | |
| 13 | MiMo-V2.5-Pro | Xiaomi | 68 |
| 14 | DeepSeek-V4-Pro-Max | DeepSeek | 68 |
| 15 | Kimi K2.6 | Moonshot AI | 67 |
| 16 | MiMo-V2.5 | Xiaomi | 66 |
| 17 | Claude Opus 4.6 | Anthropic | 65 |
| 18 | GPT-5.2 Codex | OpenAI | 64 |
| 19 | Qwen3.6 Plus | Alibaba Cloud / Qwen Team | 62 |
| 20 | GPT-5.4 mini | OpenAI | 60 |
| 21 | Qwen3.6-27B | Alibaba Cloud / Qwen Team | 59 |
| 22 | Claude Opus 4.5 | Anthropic | 59 |
| 23 | Claude Sonnet 4.6 | Anthropic | 59 |
| 24 | Muse Spark | Meta | 59 |
| 25 | MiMo-V2-Pro | Xiaomi | 57 |
| 26 | MiniMax M2.7 | MiniMax | 57 |
| 27 | DeepSeek-V4-Flash-Max | DeepSeek | 57 |
| 28 | GLM-5 | Zhipu AI | 56 |
| 29 | MAI-Code-1-Flash | Microsoft | 55 |
| 30 | Gemini 3 Pro | 54 | |
| 31 | GPT-5.1 Codex | OpenAI | 53 |
| 32 | Qwen3.5-397B-A17B | Alibaba Cloud / Qwen Team | 53 |
| 33 | Qwen3.6-35B-A3B | Alibaba Cloud / Qwen Team | 52 |
| 34 | Step-3.5-Flash | StepFun | 51 |
| 35 | Kimi K2.5 | Moonshot AI | 51 |
| 36 | Qwen3.5-122B-A10B | Alibaba Cloud / Qwen Team | 49 |
| 37 | Gemini 3 Flash | 48 | |
| 38 | DeepSeek-V3.2 (Thinking) | DeepSeek | 46 |
| 39 | DeepSeek-V3.2-Speciale | DeepSeek | 46 |
| 40 | DeepSeek-V3.2 | DeepSeek | 46 |
| 41 | GPT-5.4 nano | OpenAI | 46 |
| 42 | MAI-Thinking-1 | Microsoft | 46 |
| 43 | Qwen3.5-27B | Alibaba Cloud / Qwen Team | 42 |
| 44 | GLM-4.7 | Zhipu AI | 41 |
| 45 | Qwen3.5-35B-A3B | Alibaba Cloud / Qwen Team | 41 |
| 46 | MiMo-V2-Flash | Xiaomi | 39 |
| 47 | Qwen3-Coder 480B A35B Instruct | Alibaba Cloud / Qwen Team | 38 |
| 48 | North Mini Code 1.0 | Cohere | 36 |
| 49 | Nemotron 3 Super (120B A12B) | NVIDIA | 31 |
49 of 49 models · score normalized 0–100 where available