OCR Bench Britannica

Pairwise VLM-as-judge OCR benchmark ranking OCR models on Encyclopaedia Britannica document images with Bradley-Terry ELO scores.

6rows
eloprimary metric
2026-05-06sampled

Metadata

ID
ocr_bench_britannica
Category
Multimodal
Release
Unknown

Metrics

ELO, Win%, Wins, Losses (lower is better), Ties, ELO Low, ELO High

Latest Results

Rows are parsed from the public Hugging Face dataset-server rows API. ELO rankings come from OCR Bench pairwise VLM-as-judge comparisons.

Rank Subject ELO Model Match Provenance Sampled
1 rednote-hilab/dots.mocr 1714 — Imported 2026-05-06
2 lightonai/LightOnOCR-2-1B 1600 — Imported 2026-05-06
3 FireRedTeam/FireRed-OCR 1589 — Imported 2026-05-06
4 deepseek-ai/DeepSeek-OCR 1434 — Imported 2026-05-06
5 baidu/Qianfan-OCR 1426 — Imported 2026-05-06
6 rednote-hilab/dots.ocr 1236 — Imported 2026-05-06