Benchmark Almanac

Idefics2-8B

Idefics2-8B ranks first on MMIU.

Hugging Face Released April 9th, 2024 Family Idefics The lab's model card Compare with others
24
Benchmarks with a score
1
Ranked first, out of five or more models
2
In the top three
Hugging Face Apr 2024

Idefics2-8B

Open weights Small

No Epoch Capabilities Index score yet
  • MM-Vet 34% #139 /164
  • SEED-Bench-2-Plus 56.6% #91 /136
  • CoreCognition 59.8% #34 /119

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
MM-NIAH 11.6% #1 of 2 Official leaderboard
MMBench-CN v1.1 65.1% #123 of 178 Official leaderboard

Coding

BenchmarkScoreRankSource
ChartMimic 16.0 #14 of 14 Official leaderboard
Design2Code 81.7 #5 of 5 Official leaderboard

Vision, audio and robotics

BenchmarkScoreRankSource
MMIU 27.8% #1 of 10 Official leaderboard
CoreCognition 59.8% #34 of 119 Official leaderboard
MIRB 33% #3 of 7 Official leaderboard
MMBench 75.7% #78 of 155 Official leaderboard
MMBench-EN v1.1 72.7% #98 of 178 Official leaderboard
OmniBench 45.1% #9 of 16 Official leaderboard
A-Bench 70.5% #68 of 116 Official leaderboard
WildVision Arena 965 #6 of 10 Official leaderboard
CCBench (MMBench Chinese Culture) 37.6% #94 of 152 Official leaderboard
MMStar 49.5% #108 of 173 Official leaderboard
MuirBench 26.1% #5 of 8 Official leaderboard
SEED-Bench-2-Plus 56.6% #91 of 136 Official leaderboard
LongVideoBench 49.4% #16 of 23 Official leaderboard
MM-Vet 34% #139 of 164 Official leaderboard
VHELM 21.5 #37 of 41 Official leaderboard
CharXiv 18.2% #70 of 77 Official leaderboard
Image2Struct 12.3 #17 of 18 Official leaderboard
CharXiv (DQ) 32.8% #64 of 67 Official leaderboard
MMDU 25.4 #5 of 5 Official leaderboard

Other

BenchmarkScoreRankSource
OpenVLM Leaderboard 53.0 #116 of 173 Official leaderboard