Chameleon-30B: benchmark results

Provider: Meta. Released 2024-06-16. Access: Open.

Aggregate Rating 1256 ± 47, rank #9035 of 10489 rated models, from 418 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
OpenVLM MMT-Bench - Animated Character Recognition50Score (%)66.3
OpenVLM MMMU - Mechanical Engineering40Accuracy (%)64
OpenVLM MMT-Bench - Jigsaw Puzzle Solving30Score (%)63.3
OpenVLM MMT-Bench - Tech Engineering33.3Score (%)62.4
OpenVLM MMT-Bench - Sign Language Recognition35Score (%)60.9
OpenVLM MMT-Bench - Spot The Diff30Score (%)58.7
OpenVLM MMT-Bench - Animals Recognition90Score (%)58
OpenVLM MMT-Bench - Interactive Segmentation42.9Score (%)57.3
OpenVLM MMT-Bench - Gaze Estimation20Score (%)56.8
OpenVLM MMMU - History63.3Accuracy (%)55.1
OpenVLM MMMU - Chemistry43.3Accuracy (%)54.6
OpenVLM MMT-Bench - Navigation55Score (%)54.4

Interactive version: theaggregate.ai/model?slug=chameleon-30b · How It Works · Data refreshed daily, snapshot 2026-11-20.