General Knowledge Reasoning on MMLU-Pro
95.7AccuracyClaude Sonnet 4.5
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Claude Sonnet 4.5Input Modality=Text, LLM-as-a-Judge=GPT-4o2025.11 | 95.7 | — | — | |
| GPT-5 highInput Modality=Text, LLM-as-a-Judge=GPT-4o2025.11 | 91.4 | — | — | |
| Gemini 2.5 ProInput Modality=Text, LLM-as-a-Judge=GPT-4o2025.11 | 87.1 | — | — | |
| DeepSeek v3.22025.12 | 84.6 | — | — | |
| GLM-4.5Reported by AA Index=true2025.12 | 83.5 | — | — | |
| GLM-4.62025.12 | 83.1 | — | — | |
| INTELLECT-3Parameters=100B+2025.12 | 81.9 | — | — | |
| Sora-2 AudioInput Modality=Audio, LLM-as-a-Judge=GPT-4o2025.11 | 76.5 | — | — | |
| Std-PT.Backbone model=Deepseek-R1-Distill-Qwen-7B2026.01 | 76.02 | 1 | 1 | |
| T2Backbone model=Deepseek-R1-Distill-Qwen-7B2026.01 | 75.84 | 0.23 | 0.49 | |
| ACBackbone model=Deepseek-R1-Distill-Qwen-7B2026.01 | 75.72 | 0.47 | 2.32 | |
| ESCBackbone model=Deepseek-R1-Distill-Qwen-7B2026.01 | 75.37 | 0.38 | 1.92 | |
| DeepSeek R1 05282025.12 | 75.3 | — | — | |
| GLM-4.5-Air2025.12 | 73.9 | — | — | |
| DeepConfBackbone model=Deepseek-R1-Distill-Qwen-7B2026.01 | 73.4 | 0.59 | 3.88 | |
| T2Backbone model=Qwen3-4B2026.01 | 72.2 | 0.22 | 0.4 | |
| DeepConfBackbone model=Qwen3-4B2026.01 | 72.15 | 0.58 | 4.99 | |
| Sora-2 Last FrameInput Modality=Last Frame, LLM-as-a-Judge=GPT-4o2025.11 | 72 | — | — | |
| Std-PT.Backbone model=Qwen3-4B2026.01 | 71.98 | 1 | 1 | |
| ESCBackbone model=Qwen3-4B2026.01 | 71.9 | 0.34 | 2.23 | |
| ACBackbone model=Qwen3-4B2026.01 | 71.72 | 0.3 | 2.36 | |
| DSCBackbone model=Deepseek-R1-Distill-Qwen-7B2026.01 | 71.14 | 0.35 | 1.79 | |
| DSCBackbone model=Qwen3-4B2026.01 | 70.38 | 0.39 | 2.97 | |
| GPT-OSSParameters=120B2025.12 | 67.1 | — | — | |
| Std-PT.Backbone model=Qwen2.5-7B2026.01 | 60.18 | 1 | 1 | |
| T2Backbone model=Qwen2.5-7B2026.01 | 60.1 | 0.23 | 0.39 | |
| DeepConfBackbone model=Qwen2.5-7B2026.01 | 60.03 | 0.61 | 3.21 | |
| ACBackbone model=Qwen2.5-7B2026.01 | 59.88 | 0.32 | 2.03 | |
| ESCBackbone model=Qwen2.5-7B2026.01 | 59.88 | 0.28 | 1.78 | |
| DSCBackbone model=Qwen2.5-7B2026.01 | 59.59 | 0.4 | 3.74 | |
| ESCBackbone model=Llama3.1-8B2026.01 | 52.63 | 0.35 | 2.16 | |
| ACBackbone model=Llama3.1-8B2026.01 | 52.58 | 0.31 | 3.17 | |
| Std-PT.Backbone model=Llama3.1-8B2026.01 | 52.54 | 1 | 1 | |
| T2Backbone model=Llama3.1-8B2026.01 | 52.37 | 0.22 | 0.42 | |
| DSCBackbone model=Llama3.1-8B2026.01 | 52 | 0.26 | 2.83 | |
| DeepConfBackbone model=Llama3.1-8B2026.01 | 49.86 | 0.63 | 4.16 |