ModelRadar
/

Jina OCR V1

Jina AI๐Ÿ‡ฉ๐Ÿ‡ช DEMajor labยท Sep 1, 2026

TextVisionopen weights

Jina OCR V1 is a document parsing model built on DeepSeek-OCR that uses FastMTP speculative decoding for faster page transcription. It scores 91.14 on OmniDocBench v1.6 and 83.4 on olmOCR-Bench.

Summary of the model card

License
cc-by-nc-4.0
Modalities
text, image โ†’ text

1 variant

StandardHF

Sources

HuggingFacePaper
Deutsch

Jina OCR V1 ist ein Modell zum Parsen von Dokumenten, das auf DeepSeek-OCR aufbaut und per FastMTP-Spekulativdekodierung schneller arbeitet. Es erreicht 91.14 auf OmniDocBench v1.6 und 83.4 auf olmOCR-Bench.

Provenance

Detected
Sep 27, 2026
First source
huggingface
Description
Summary of the model card
Editorially reviewed
โ€”