General Knowledge Exam on MMLU-Pro (Accuracy)
0.8645MMLU-Pro AccuracyGemini-2.5 (Pro)
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-2.5 (Pro)Model Tier=Pro2026.01 | 0.8645 | |
| Qwen3-VL (Thinking)Thinking Mode=true, Parameters=235B-A22B2026.01 | 0.8375 | |
| Seed-1.5-VL (Thinking)Thinking Mode=true2026.01 | 0.8339 | |
| GLM-4.6VParameters=106B-A12B2026.01 | 0.7996 | |
| STEP3-VL-10B (PaCoRe)Reasoning Strategy=PaCoRe, Parameters=10B2026.01 | 0.7709 | |
| STEP3-VL-10B (SeRe)Reasoning Strategy=SeRe, Parameters=10B2026.01 | 0.7602 |