Lukaszl/clearocr-benchmark-fiszki91-v1-results
OCR Bench Results: clearocr-benchmark-fiszki91-v1 VLM-as-judge pairwise evaluation of OCR models. Rankings depend on document type — there is no single best OCR model. Leaderboard Rank Model Params ELO 95% CI Wins Losses Ties Win% 1 clearocr.com/clearocr-api 1777 1734–1824 316 47 0 87% 2 lightonai/LightOnOCR-2-1B 1B 1567 1537–1603 222 141 0 61% 3 deepseek-ai/DeepSeek-OCR 4B 1412 1378–1444 137 225 0 38% 4 FireRedTeam/FireRed-OCR 2.1B 1380… See the full description on the dataset page: https://huggingface.co/datasets/Lukaszl/clearocr-benchmark-fiszki91-v1-results.
OCR Bench Results: clearocr-benchmark-fiszki91-v1
VLM-as-judge pairwise evaluation of OCR models. Rankings depend on document type — there is no single best OCR model.
Leaderboard
Details
- Task: OCR (Optical Character Recognition)
- Language: Polish
- Document type: scanned newspaper clippings with varying print quality, scan condition, and occasional handwritten annotations
- Original dataset: `Zombely/fiszki-ocr-test-A`
- Source dataset: `Lukaszl/clearocr-benchmark-fiszki91-v1`
- Judge: Qwen3.5-35B-A3B
- Comparisons: 907
- Method: Bradley-Terry MLE with bootstrap 95% CIs
About clearOCR
clearOCR is an OCR API for extracting text from PDFs, scans and document images, with a strong focus on Polish and English documents.
New accounts currently receive:
- 1,000 free single-image OCR runs
- valid for 30 days
API access is available via the clearOCR website:
https://clearocr.com
Configs
load_dataset("Lukaszl/clearocr-benchmark-fiszki91-v1-results")— leaderboard tableload_dataset("Lukaszl/clearocr-benchmark-fiszki91-v1-results", name="comparisons")— full pairwise comparison logload_dataset("Lukaszl/clearocr-benchmark-fiszki91-v1-results", name="metadata")— evaluation run history
Generated by [ocr-bench](https://github.com/davanstrien/ocr-bench)
