PandaGPT-13B: benchmark results

Provider: UCSD. Released 2023-05-25.

Aggregate Rating 1209 ± 52, rank #9366 of 10489 rated models, from 422 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
OpenVLM MMT-Bench - Gaze Estimation40Score (%)99
OpenVLM MMT-Bench - Multiple View Image Understanding50Score (%)98.1
OpenVLM MMT-Bench - Font Recognition35Score (%)88.1
OpenVLM MMT-Bench - Order Hallucination50Score (%)81.8
OpenVLM MMT-Bench - Micro Expression Recognition35Score (%)79.4
OpenVLM MMT-Bench - Web Shopping45Score (%)78.9
OpenVLM MMT-Bench - Artwork Emotion Recognition55Score (%)74.8
OpenVLM MMT-Bench - Image Quality Assessment50Score (%)72.6
OpenVLM MMT-Bench - GUI Install45Score (%)71.6
OpenVLM MMT-Bench - Action Quality Assessment25Score (%)69.7
OpenVLM MMT-Bench - Handwritten Retrieval30Score (%)67.2
AVI-Bench - Visual-Referenced Audio Retrieval23.7Task score (%) on visual-referenced audio retrieval (264 que64

Interactive version: theaggregate.ai/model?slug=pandagpt-13b · How It Works · Data refreshed daily, snapshot 2026-11-20.