Text Math Reasoning on TAL-SCQ5K-CN
98.8ScoreGPT-5.4
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5.4Parameter Count=≈1.5T2026.06 | 98.8 | |
| Qwen3.5-PlusParameter Count=397B2026.06 | 96.9 | |
| Opus-4.6Parameter Count=≈800B2026.06 | 95.1 | |
| Qwen3-VL-32BDecoding Mode=Thinking, Model Scale=32B2026.06 | 90.2 | |
| K2.5Parameter Count=≈1T2026.06 | 88.7 | |
| Qwen3-VL-32BDecoding Mode=Instruct, Model Scale=32B2026.06 | 81.2 | |
| Yuvion VL-32BDecoding Mode=Reasoning, Model Scale=32B2026.06 | 80.1 | |
| Qwen3-VL-8BDecoding Mode=Thinking, Model Scale=8B2026.06 | 78.3 | |
| Qwen3-VL-8BDecoding Mode=Instruct, Model Scale=8B2026.06 | 75.4 | |
| Yuvion VL-8BDecoding Mode=Reasoning, Model Scale=8B2026.06 | 75 | |
| Yuvion VL-32BDecoding Mode=Instruct, Model Scale=32B2026.06 | 74.7 | |
| Yuvion VL-8BDecoding Mode=Instruct, Model Scale=8B2026.06 | 70.9 |