Benchmark Almanac

Qwen-VL-Max

Qwen-VL-Max ranks first on MME.

Released 18 Jan 2024
20
Benchmarks with a score
1
Ranked first, out of five or more models
3
In the top three
Unknown lab Jan 2024

Qwen-VL-Max

No Epoch Capabilities Index score yet
  • OlympicArena 21.4% #12 /25
  • SpatialViz-Bench 32% #7 /8
  • OlympiadBench 10.1% #3 /7

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

BenchmarkScoreRankSource
MMBench-CN v1.1 73.7% #83 of 178 Official leaderboard

Math and science

BenchmarkScoreRankSource
OlympiadBench 10.1% #3 of 7 Official leaderboard
OlympicArena 21.4% #12 of 25 Official leaderboard

Agents and tools

BenchmarkScoreRankSource
SnakeBench 28 #50 of 183 Official leaderboard

Vision, audio and robotics

BenchmarkScoreRankSource
MME 1790.0 #1 of 16 Official leaderboard
MathVerse 25.3% #2 of 9 Official leaderboard
CCBench (MMBench Chinese Culture) 63.5% #34 of 152 Official leaderboard
ERQA 39.3% #6 of 17 Official leaderboard
MMBench 77.6% #71 of 155 Official leaderboard
MMBench-EN v1.1 75.4% #87 of 178 Official leaderboard
BLINK 41.9% #6 of 11 Official leaderboard
MTVQA-Test 21.1% #10 of 18 Official leaderboard
MMMU 51.4% #79 of 136 Official leaderboard
MMStar 49.5% #107 of 173 Official leaderboard
Vision Arena 1,185 #87 of 136 Official leaderboard
CharXiv 24.7% #57 of 77 Official leaderboard
SpatialViz-Bench 32% #7 of 8 Epoch AI
Video-MME 51.3% #23 of 26 Official leaderboard
MATH-Vision (MathVision) 15.6% #96 of 105 Official leaderboard

Other

BenchmarkScoreRankSource
OpenVLM Leaderboard 58.4 #91 of 173 Official leaderboard