Unlimited OCR vs Pixtral
Pick Unlimited OCR to parse long reports and books into Markdown with bounding boxes, for example as bulk PDF to text for RAG. Pick Pixtral only if you already run on the Mistral API and need question answering over charts and documents. Pixtral is deprecated and no longer maintained, so it is a poor base for a new OCR project.
Unlimited OCRbaidu/unlimited-ocr | Pixtralmistral/pixtral | |
|---|---|---|
| Type | Open source | Vision model |
| Price | 2 credits / page | 2 credits / page |
| Per 1,000 pages | $2.00 | $2.00 |
| Typical speed | 1.2 s / page | 3.0 s / page |
| Best for | Books and very long documents | Questions about document images |
| Inputs | PDF, images | images |
| Languages | 30+ languages | Multilingual |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | No |
| Handwriting | No | No |
| Math | ✓ Yes | No |
| Bounding boxes | ✓ Yes | No |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | ✓ Yes | ✓ Yes |
| On OCRRouter | Available | Available |
Prices are what you pay on OCRRouter. Speed is the provider's typical figure, not a benchmark we ran.
Which to choose
Choose Unlimited OCR if
- It parses 40+ pages in a single forward pass, so a whole report or book goes through in one request.
- It returns Markdown, plain text and bounding boxes, with block types alongside the text.
- It is MIT licensed and maintained, so you can use it commercially with no copyleft.
Choose Pixtral if
- It scores 90.7 ANLS on DocVQA per Mistral, which suits question answering over documents and charts.
- It has a 128K token context, so one request can carry many images.
- The Mistral API launch price was $0.15 per million input and output tokens, and the weights are Apache-2.0 if you self-host.
Key differences
| Unlimited OCR | Pixtral | |
|---|---|---|
| Main job | Multi-page document parser built from DeepSeek-OCR, 3B Mixture-of-Experts | General vision language model, 12.4B parameters (12B decoder plus 400M vision encoder) |
| Licence | MIT | Apache-2.0 |
| How you run it | Self-host with vLLM, SGLang or transformers, or use the hosted endpoint in Baidu Cloud's OCR product. Needs an NVIDIA GPU with CUDA 12.9 or 13.0 | Self-host the open weights, or use the Mistral API as pixtral-12b-2409 |
| Price | Free to self-host, you pay for compute. Hosted price Not published | $0.15 per million input and output tokens at API launch. Free to self-host, you pay for compute |
| Output formats | Markdown, plain text, bounding boxes | Plain text, Markdown, JSON |
| Context and length | Context capped at 32,768 tokens. Multi-page input only works in base mode | 128K token context, many images per request |
| Known failures | Text repetition and looping (issue #55), hallucinated text (issue #58), bold and italic styling mostly lost, 12 GB GPUs run out of memory with version 2 under SGLang | Deprecated and unmaintained. Reported accuracy drops on handwriting, small fonts and non-Latin scripts. Q4 quantization degrades small text OCR |
Sources
- Hugging Face model card
- GitHub repo
- MarkTechPost on Unlimited OCR 2
- Mistral announcement
- Mistral Sept 2024 pricing
What each one is
Unlimited OCR
Open-source model from Baidu that reads hundreds of pages in one pass with flat memory use.
Unlimited OCR detailsPixtral
Mistral's open-weight vision models, Pixtral 12B and Pixtral Large. Mistral's API no longer serves them, so they run self-hosted.
Pixtral detailsTry both on your own file
Run the file once with each model and compare the output.
Use one, fall back to the other
{"model": "baidu/unlimited-ocr","fallbacks": ["mistral/pixtral"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Unlimited OCR alternativesPixtral alternativesBest OCR models in 2026