# baidu/Unlimited-OCR

Unlimited OCR Works: Welcome the Era of One-shot Long-horizon Parsing.

Repository: https://github.com/baidu/Unlimited-OCR
Canonical: https://ross.abutalabs.com/products/unlimited-ocr
Language: Python
License: MIT
License Family: permissive
Last push: 2026-07-29T04:37:03+00:00

## Health v2 (maintenance only)
Score: 56/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 95, release rhythm 35, longevity 5
- inputs: {"age_days": 76, "days_push": 35, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 24569, forks 2543 (observed 2026-08-28T04:11:37.172142+00:00)

## What it is
Baidu's Unlimited-OCR is an open vision-language OCR model for one-shot long-horizon document parsing, extending DeepSeek-OCR. It provides inference via Hugging Face Transformers and vLLM on NVIDIA GPUs, with weights on Hugging Face and ModelScope.

## Use cases
- extract text from long pdf documents
- parse scanned documents with a vision-language model
- run ocr on gpu with vllm
- convert document images to structured text
- batch document parsing for rag pipelines

## When to choose
- you need ocr over very long or multi-page documents in one pass
- you want an open model deployable with transformers or vllm
- you need document parsing as input for downstream llm workflows

## When to avoid
- you need lightweight cpu-only ocr like tesseract
- you need a hosted ocr api without gpu infrastructure
- you need handwriting or niche-language recognition without evaluation

## Facets
- artifact type: library
- maturity: active
- function: ocr, machine-learning, llm-inference, pdf
- domain: computer-vision, deep-learning, pdf
- platform: python
- tags: vision-language-model, document-parsing, long-context, huggingface, vllm, natural-language-processing, gpu, linux

## Member repositories
- baidu/Unlimited-OCR (main) score 56

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:37.172142+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:56:12.105119+00:00, confidence not recorded.
  - readme: https://github.com/baidu/Unlimited-OCR (fetched 2026-08-28T04:11:37.172142+00:00, sha 5f77b9450d49)
- Data as of 2026-08-30T08:39:29.467469+00:00.