# MiniCPM-OCR Jev model sizes and sources

> MiniCPM-OCR Jev is OpenBMB's 2B model for scoring candidate OCR readings against document regions. It returns a score from 0 to 1.

Updated 2026-11-04 (Australia/Adelaide)

By OpenBMB · First Hugging Face upload: 2026-10-23 · OPEN

This 2B specialist scores whether a candidate OCR reading matches a document region. It handles text, formulas and tables, returning a score from 0 to 1. Qwen3.5-2B supplies the backbone; a binary classification head replaces its language head. Scores can help select among OCR engines or route uncertain regions to a person.

## Facts

- Size: 2B
- Base: Qwen3.5-2B
- Licence: Apache-2.0
- Access: Open weights in BF16, AWQ, GPTQ and GGUF; a free API on ModelBest's platform

## Claims and sources

- **VENDOR-REPORTED**: OpenBMB reports decision AUROC ×100 of 93.6 on its JevOCRBench, against 90.1 for a row named JEV-35B-A3B and 93.8 for Qwen3.5-397B-A17B. The card does not identify JEV-35B-A3B further. [source](<https://huggingface.co/openbmb/MiniCPM-OCR-Jev>)
- **VENDOR-REPORTED**: Across four OCR parsers on the same 200 samples with manually labelled readings, OpenBMB reports AUROC ×100 from 91.9 to 93.4. These are its own parser and sample combinations. [source](<https://huggingface.co/openbmb/MiniCPM-OCR-Jev>)

## Links

- [Source: Weights and model card](<https://huggingface.co/openbmb/MiniCPM-OCR-Jev>)
- [Source: Code and deployment](<https://github.com/OpenBMB/MiniCPM-OCR>)
- [Source: JevOCRBench](<https://huggingface.co/datasets/openbmb/JevOCRBench>)
- [Source: Hugging Face upload metadata](<https://huggingface.co/api/models/openbmb/MiniCPM-OCR-Jev>)

## Related projects

No related projects are listed in the registry.