| Model | MM-K12 | OlympiadBench | MathVista | MathVerse | MathVision |
| MM-Policy | 33.92 | 15.41 | 62.93 | 42.99 | 21.74 |
| +MM-PRM | 42.80 | 24.00 | 67.60 | 46.27 | 27.11 |
| +8.88 | +8.59 | +4.67 | +3.28 | +5.37 |
| InternVL2.5-8B | 27.01 | 5.23 | 56.43 | 36.26 | 10.04 |
| +MM-PRM | 37.80 | 15.33 | 63.50 | 42.56 | 19.41 |
| +10.79 | +10.10 | +7.07 | +6.30 | +9.37 |
| InternVL2.5-26B | 28.01 | 14.46 | 60.02 | 37.83 | 20.76 |
| +MM-PRM | 38.00 | 24.67 | 64.50 | 44.19 | 25.63 |
| +9.99 | +10.21 | +4.25 | +6.36 | +4.87 |
| InternVL2.5-38B | 40.34 | 29.57 | 68.32 | 47.94 | 29.70 |
| +MM-PRM | 52.40 | 32.67 | 71.10 | 52.61 | 32.99 |
| +12.06 | +3.10 | +2.78 | +4.67 | +3.29 |
| InternVL2.5-78B | 42.24 | 30.98 | 69.48 | 50.18 | 31.50 |
| +MM-PRM | 48.80 | 34.67 | 73.20 | 54.47 | 33.26 |
| +6.56 | +3.69 | +3.72 | +4.29 | +1.76 |
