Reasoning on AGI Eval EN
89.4AccuracyQwen 3 VL 32B Instruct
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen 3 VL 32B InstructParameters=32B2025.12 | 89.4 | |
| DictaLM 3.0 24B-ThinkParameters=24B, Variant=Thinking2026.02 | 82.93 | |
| Qwen 3 32BThinking=No, Parameters=32B2025.12 | 82.4 | |
| Olmo 3.1 32B InstructStage=Final Instruct 3.12025.12 | 79.5 | |
| Olmo 3.1 32B InstructStage=DPO2025.12 | 79.4 | |
| Qwen 2.5 32BParameters=32B2025.12 | 78.9 | |
| Gemma 3 27BParameters=27B2025.12 | 76.9 | |
| Gemma 3 27BParameters=27B2026.02 | 76.9 | |
| Mistral Small 3.12026.02 | 75.87 | |
| Gemma 3 12BParameters=12B2026.02 | 73.91 | |
| DictaLM 3.0 12B-InstParameters=12B, Variant=Instruct2026.02 | 73.75 | |
| Olmo 3.1 32B InstructStage=SFT2025.12 | 71.7 | |
| Gemma 2 27BParameters=27B2025.12 | 70.9 | |
| OLMo 2 32BParameters=32B2025.12 | 68.4 | |
| Apertus 70BParameters=70B2025.12 | 61.6 |