Codesota · OCR · Benchmarks · MME-VideoOCRHome/OCR/Benchmarks/MME-VideoOCR
NTU

MME-VideoOCR.

Video OCR benchmark with 1,464 videos and 2,000 QA pairs across 25 tasks.

View on AlphaXiv
§ 01 · Total Accuracy

Total Accuracy.

Overall accuracy across all video OCR tasks

Higher is better

#ModelScoreSource
gemini-25-pro
Non-API entry from src
73.7%src
2
qwen25-vl-72b
Non-API entry from src
69%src
3
internvl3-78b
Non-API entry from src
67.2%src
4
gpt-4o
Non-API entry from src
66.4%src
5
gemini-15-pro
Non-API entry from src
64.9%src
6
qwen25-vl-32b
Non-API entry from src
61%src
§ Related · Explore

More OCR content.

Verified Model Reviews
Comparisons & Guides
View all OCR benchmarks → Back to All Benchmarks