Benchmark Almanac

mPLUG-Owl2

mPLUG-Owl2 places in the top three on NaturalBench.

Alibaba Group Released November 2nd, 2023 Family mPLUG-Owl The lab's model card Compare with others
21
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
Alibaba Group Nov 2023

mPLUG-Owl2

Open weights Small

No Epoch Capabilities Index score yet
  • SEED-Bench 57.8% #6 /9
  • MM-Vet 35.7% #135 /164
  • SEED-Bench-2-Plus 44.4% #120 /136

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
AMBER 84.0 #1 of 2 Official leaderboard
MMBench-CN v1.1 58% #149 of 178 Official leaderboard

Agents and tools

BenchmarkScoreRankSource
VisualWebBench 18.6% #5 of 7 Official leaderboard

Professional work

BenchmarkScoreRankSource
GMAI-MMBench 36.2% #9 of 13 Official leaderboard

Vision, audio and robotics

BenchmarkScoreRankSource
NaturalBench 17.4% #2 of 6 Official leaderboard
CoreCognition 53.7% #73 of 119 Official leaderboard
SEED-Bench 57.8% #6 of 9 Official leaderboard
CCBench (MMBench Chinese Culture) 31% #107 of 152 Official leaderboard
MMT-Bench 52% #11 of 15 Official leaderboard
MM-Vet 35.7% #135 of 164 Official leaderboard
MME 1450.2 #14 of 17 Official leaderboard
MMBench 66% #129 of 155 Official leaderboard
MMBench-EN v1.1 63.5% #152 of 178 Official leaderboard
SEED-Bench-2-Plus 44.4% #120 of 136 Official leaderboard
OCRBench 36.6% #10 of 11 Official leaderboard
MMStar 34.8% #162 of 173 Official leaderboard
LongVideoBench 39.4% #22 of 23 Official leaderboard
MMMU 32.7% #133 of 136 Official leaderboard
CMMMU 22.2% #9 of 9 Official leaderboard
Q-Bench-Video 43.4% #3 of 3 Official leaderboard

Other

BenchmarkScoreRankSource
OpenVLM Leaderboard 37.8 #164 of 173 Official leaderboard