Skip to content

Benchmark

Video-MME-v2 - Reasoning Coherence

Vision benchmarks measure how well models understand images. Scores are percentages.

Models measured
20
Scores recorded
78
Sources
1
IQ weight
1.71

Score distribution

22%
Mean
22.4%
Median
9%
Std dev
11.1%
10th percentile
35.4%
90th percentile
6.7%
Min
6.7%47%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.71 for chat models, measured over 20 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Gemini 3 Pro Previeww. sub @ 1fps· Video-MME-v247%
2Doubao Seed 2.0 Prow. sub @ 1fps· Video-MME-v242.5%
3Kimi K2.5w. sub @ 512· Video-MME-v240.9%
4Qwen3.5 397B A17Bw. sub @ 512, think· Video-MME-v239.4%
5Gemini 3 Flashw. sub @ 1fps· Video-MME-v238.2%
6MiMo V2 Omniw. sub @ 1fps· Video-MME-v236.6%
7GPT 5w. sub @ 50· Video-MME-v236.3%
8Kimi K2.5w. sub @ 64· Video-MME-v236.2%
9Gemini 3 Pro Previewwo sub @ 1fps· Video-MME-v235.4%
10Qwen3.5 397B A17Bw. sub @ 64, think· Video-MME-v232.7%
11Qwen3.5 27Bw. sub @ 64, think· Video-MME-v231.7%
12Doubao Seed 2.0 Prowo sub @ 1fps· Video-MME-v231.6%
13Qwen3.5 27Bw. sub @ 512, think· Video-MME-v231.3%
14Kimi K2.5wo sub @ 512· Video-MME-v231%
15Gemma 4 31B ITw. sub @ 60· Video-MME-v230.9%
16Gemini 3 Flashwo sub @ 1fps· Video-MME-v228%
17Qwen3 VL 235Bw. sub @ 512, think· Video-MME-v227.8%
18Qwen3.5 35B A3Bw. sub @ 64, think· Video-MME-v227.5%
19Qwen3.5 122B-A10Bw. sub @ 64, think· Video-MME-v227.3%
20Qwen3 VL 235Bw. sub @ 64, think· Video-MME-v227%
21Qwen3.5 Omni Plusw. sub @ 512· Video-MME-v227%
22Qwen3.5 397B A17Bwo sub @ 512, think· Video-MME-v226.9%
23Qwen3.5 9Bw. sub @ 512, think· Video-MME-v226.8%
24Qwen3.5 122B-A10Bw. sub @ 512, think· Video-MME-v226.5%
25Qwen3.5 Omni Plusw. sub @ 64· Video-MME-v226%
26Qwen3.5 397B A17Bw. sub @ 64, instruct· Video-MME-v225.7%
27MiMo V2 Omniwo sub @ 1fps· Video-MME-v225.4%
28Qwen3 VL 235Bw. sub @ 64, instruct· Video-MME-v225%
29Kimi K2.5wo sub @ 64· Video-MME-v224.9%
30Qwen3.5 27Bw. sub @ 64, instruct· Video-MME-v224.8%
31Qwen3.5 35B A3Bw. sub @ 512, think· Video-MME-v224.7%
32GLM 4.5Vw. sub @ 64, think· Video-MME-v224.3%
33Qwen2.5 VL 72B Instructw. sub @ 64, instruct· Video-MME-v224.3%
34GPT 5wo sub @ 50· Video-MME-v223.2%
35qwen3-vl-30bw. sub @ 512, think· Video-MME-v222.8%
36Qwen3.5 35B A3Bw. sub @ 64, instruct· Video-MME-v222.7%
37qwen3-vl-30bw. sub @ 64, think· Video-MME-v222.6%
38internvl3-5-241b-a28bw. sub @ 64, instruct· Video-MME-v222.5%
39Qwen3.5 122B-A10Bw. sub @ 64, instruct· Video-MME-v222.4%
40Qwen3.5 9Bw. sub @ 64, think· Video-MME-v222.3%
41Qwen3.5 4Bw. sub @ 64, think· Video-MME-v221.4%
42Qwen3.5 9Bw. sub @ 64, instruct· Video-MME-v221.2%
43Qwen: Qwen3 VL 8B Thinkingw. sub @ 64, think· Video-MME-v220.5%
44Qwen3.5 397B A17Bwo sub @ 64, think· Video-MME-v220.4%
45Qwen: Qwen3 VL 8B Thinkingw. sub @ 512, think· Video-MME-v220.3%
46Qwen: Qwen3 VL 8B Thinkingw. sub @ 64, instruct· Video-MME-v220.3%
47qwen3-vl-30bw. sub @ 64, instruct· Video-MME-v219%
48Gemma 4 31B ITwo sub @ 60· Video-MME-v218.3%
49Qwen3.5 Omni Pluswo sub @ 64· Video-MME-v217.2%
50Qwen3.5 9Bwo sub @ 512, think· Video-MME-v216.8%

Showing 1–50 of 78 results.