How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites

Open paper
Benchmark Model Rank Results
visual-question-answering-on-mm-vetInternVL 1.5#22GPT-4 score: 62.8Params: 26B
visual-question-answering-on-mm-vetInternVL 1.2#50GPT-4 score: 48.9Params: 40B
visual-question-answering-on-mm-vet-v2InternVL-Chat-V1-5#9GPT-4 score: 51.5±0.2

Rank counts only results with a code link.