| 1 | Claude Opus 5.5 | Anthropic | Proprietary | 87.7% | |
| 2 | GPT-6 Astra | OpenAI | Proprietary | 86.9% | |
| 3 | Gemini 3.8 Flash | Google | Proprietary | 85.6% | |
| 4 | Gemini 3.7 Flash | Google · Closed | 85.5% | | |
| 5 | Claude Opus 5 | Anthropic · Closed | 84.7% | | |
| 6 | Gemini 3.5 Flash | Google · Closed | 84.3% | | |
| 7 | GPT-5.6 Sol | OpenAI · Closed | 83.4% | | |
| 8 | GPT-6 Sol | OpenAI · Closed | 83.3% | | |
| 9 | Gemini 3.6 Flash | Google · Closed | 83.2% | | |
| 10 | Qwen3.8 Max Preview | Alibaba · Closed | 82.8% | | |
| 11 | Gemini 3.1 Pro | Google · Closed | 82.4% | | |
| 12 | GPT-5.6 Terra | OpenAI · Closed | 80.7% | | |
| 13 | Kimi K3 | Moonshot AI · Closed | 80.5% | | |
| 14 | Qwen3.7 Plus | Alibaba · Closed | 80.5% | | |
| 15 | Muse Spark | Meta · Closed | 80.5% | | |
| 16 | Grok 4.5 | xAI · Closed | 80.4% | | |
| 17 | Gemini 3 Pro | Google · Closed | 80.2% | | |
| 18 | GPT-5.5 | OpenAI · Closed | 79.9% | | |
| 19 | Qwen3.8-Flash-Next | Alibaba · Open weight | 79.8% | | |
| 20 | Kimi K2.6 | Moonshot AI · Open weight | 79.4% | | |
| 21 | Apodex 1.1 | Apodex · Closed | 79.2% | | |
| 22 | Apodex 1.1 Mini | Apodex · Open weight | 79.2% | | |
| 23 | Ling 3.0 Flash VL | InclusionAI · Open weight | 79.0% | | |
| 24 | Gemini 3.5 Flash-Lite | Google · Closed | 79.0% | | |
| 25 | Claude Opus 4.7 (Adaptive) | Anthropic · Closed | 78.8% | | |
| 26 | GPT-5.6 Luna | OpenAI · Closed | 78.6% | | |
| 27 | MiniMax M3 | MiniMax · Open weight | 78.6% | | |
| 28 | Gemini 3 Flash | Google · Closed | 78.6% | | |
| 29 | GPT-5.3 Codex | OpenAI · Closed | 78.5% | | |
| 30 | GPT-5.3-Codex-Spark | OpenAI · Closed | 78.5% | | |
| 31 | GPT-5.4 | OpenAI · Closed | 78.4% | | |
| 32 | Grok 4.3 | xAI · Closed | 78.1% | | |
| 33 | Qwen3.6 Plus | Alibaba · Closed | 78.0% | | |
| 34 | Claude Sonnet 5 | Anthropic · Closed | 77.3% | | |
| 35 | DeepSeek V4.1 Flash | DeepSeek · Open weight | 77.0% | | |
| 36 | Step 5 Preview | StepFun · Closed | 76.4% | | |
| 37 | Claude Opus 4.7 | Anthropic · Closed | 76.4% | | |
| 38 | GPT-5.2-Codex | OpenAI · Closed | 76.3% | | |
| 39 | Qwen3.8-27B | Alibaba · Open weight | 76.3% | | |
| 40 | GPT-6 Luna | OpenAI · Closed | 75.5% | | |
| 41 | GPT-5.1 | OpenAI · Closed | 75.5% | | |
| 42 | Kimi K2.5 | Moonshot AI · Open weight | 75.4% | | |
| 43 | Kimi K2.5 (Reasoning) | Moonshot AI · Closed | 75.4% | | |
| 44 | Claude Opus 4.6 (Adaptive) | Anthropic · Closed | 75.4% | | |
| 45 | Step 3.7 Flash | StepFun · Open weight | 75.3% | | |
| 46 | Qwen3.5-27B | Alibaba · Open weight | 75.0% | | |
| 47 | Qwen3.5-122B-A10B | Alibaba · Open weight | 75.0% | | |
| 48 | Qwen3.6-35B-A3B | Alibaba · Open weight | 75.0% | | |
| 49 | Gemini 2.5 Pro | Google · Closed | 74.9% | | |
| 50 | Qwen3.6-27B | Alibaba · Open weight | 74.6% | | |
| 51 | Muse Glimmer 30B | Meta · Open weight | 74.3% | | |
| 52 | GPT-5 (medium) | OpenAI · Closed | 74.3% | | |
| 53 | GPT-5 (high) | OpenAI · Closed | 74.2% | | |
| 54 | Inkling-Small | Thinking Machines Lab · Open weight | 74.0% | | |
| 55 | Claude Opus 4.5 Thinking | Anthropic · Closed | 74.0% | | |
| 56 | Inkling | Thinking Machines Lab · Open weight | 73.5% | | |
| 57 | Gemma 4 31B | Google · Open weight | 73.4% | | |
| 58 | GPT-5.4 mini | OpenAI · Closed | 73.3% | | |
| 59 | GLM-5V-Turbo | Z.AI · Closed | 72.8% | | |
| 60 | Qwen3.5-35B-A3B | Alibaba · Open weight | 72.7% | | |
| 61 | Claude Opus 4.6 | Anthropic · Closed | 72.5% | | |
| 62 | GPT-5.1-Codex | OpenAI · Closed | 72.5% | | |
| 63 | GPT-5.1-Codex-Max | OpenAI · Closed | 72.5% | | |
| 64 | Claude Opus 4.5 | Anthropic · Closed | 71.2% | | |
| 65 | Claude Sonnet 4.6 | Anthropic · Closed | 70.6% | | |
| 66 | o3 | OpenAI · Closed | 70.1% | | |
| 67 | GPT-5 mini | OpenAI · Closed | 70.1% | | |
| 68 | MiMo-V2-Omni | Xiaomi · Closed | 69.9% | | |
| 69 | Gemma 4 12B | Google · Open weight | 69.7% | | |
| 70 | Gemma 4 26B A4B | Google · Open weight | 69.2% | | |
| 71 | Grok 4 | xAI · Closed | 68.8% | | |
| 72 | Claude 4.1 Opus Thinking | Anthropic · Closed | 67.9% | | |
| 73 | Gemini 2.5 Flash | Google · Closed | 65.5% | | |
| 74 | GPT-5.4 nano | OpenAI · Closed | 65.4% | | |
| 75 | Mistral Medium 3.5 128B | Mistral · Open weight | 64.9% | | |
| 76 | Grok 4.1 Fast (Reasoning) | xAI · Closed | 63.3% | | |
| 77 | Command A+ | Cohere · Open weight | 63.2% | | |
| 78 | Claude 4 Sonnet | Anthropic · Closed | 62.4% | | |
| 79 | Llama 4 Maverick | Meta · Open weight | 62.1% | | |
| 80 | Grok 4 Fast (Reasoning) | xAI · Closed | 61.8% | | |
| 81 | GPT-4.1 | OpenAI · Closed | 61.2% | | |
| 82 | GPT-5 nano | OpenAI · Closed | 61.0% | | |
| 83 | Qwen3-Omni-30B-A3B-Thinking | Alibaba · Open weight | 60.2% | | |
| 84 | GPT-4.1 mini | OpenAI · Closed | 58.7% | | |
| 85 | Mistral Small 4 | Mistral · Open weight | 56.8% | | |
| 86 | Mistral Small 4 (Reasoning) | Mistral · Open weight | 56.8% | | |
| 87 | Mistral Large 3 | Mistral · Closed | 55.7% | | |
| 88 | Qwen3-Omni-30B-A3B-Instruct | Alibaba · Open weight | 55.5% | | |
| 89 | Gemini 1.5 Pro | Google · Closed | 55.0% | | |
| 90 | Nemotron 3 Nano Omni 30B A3B | NVIDIA · Open weight | 53.2% | | |
| 91 | Mistral Medium 3 | Mistral · Closed | 53.0% | | |
| 92 | Llama 4 Scout | Meta · Open weight | 52.9% | | |
| 93 | Qwen3.5 397B | Alibaba · Open weight | 52.7% | | |
| 94 | Qwen3.5 397B (Reasoning) | Alibaba · Open weight | 52.7% | | |
| 95 | Gemma 4 E4B | Google · Open weight | 51.4% | | |
| 96 | Ministral 3 14B (Reasoning) | Mistral · Open weight | 49.8% | | |
| 97 | Ministral 3 14B | Mistral · Open weight | 49.8% | | |
| 98 | Grok 4.1 Fast | xAI · Closed | 48.4% | | |
| 99 | Gemma 3 27B | Google · Open weight | 48.0% | | |
| 100 | Ministral 3 8B (Reasoning) | Mistral · Open weight | 46.0% | | |
| 101 | Ministral 3 8B | Mistral · Open weight | 46.0% | | |
| 102 | Gemma 4 E2B | Google · Open weight | 44.6% | | |
| 103 | Nova Pro | Amazon · Closed | 44.3% | | |
| 104 | GPT-4o mini | OpenAI · Closed | 41.5% | | |
| 105 | GPT-4.1 nano | OpenAI · Closed | 40.1% | | |
| 106 | Ministral 3 3B (Reasoning) | Mistral · Open weight | 38.1% | | |
| 107 | Ministral 3 3B | Mistral · Open weight | 38.1% | | |
| 108 | Claude 3 Haiku | Anthropic · Closed | 30.8% | | |
| 109 | LFM2.5-VL-1.6B-Extract | LiquidAI · Open weight | 26.5% | | |
| 110 | Phi-4 Multimodal Instruct | Microsoft · Open weight | 14.5% | | |