OCRBench V2

A native OCR benchmark for reading text from images across multilingual scripts, low-quality scans, handwriting, structured layouts, charts, and screenshots.

ModelsModels with a result.6
Top resultBest result on this benchmark.—
Top-3 spreadPoints from first to third.—
YearYear of release.2025

Results

6 results
#ModelResultThe score from the source.
—Qwen3.8 Max74.2%
—Qwen3.7 Plus70.7%
—LFM3-VL-2B63.5%