Kraken vs Pixtral
Pick Kraken for historical manuscripts, early prints and right-to-left scripts such as Arabic and Hebrew, where you can train layout and recognition on your own material. Pick Pixtral only if you already run on the Mistral API and want question answering over charts and documents. Mistral marks Pixtral 12B as deprecated, so it is a poor base for a new OCR project.
Krakenkraken/kraken | Pixtralmistral/pixtral | |
|---|---|---|
| Type | Open source | Vision model |
| Price | 1 credit / page | 2 credits / page |
| Per 1,000 pages | $1.00 | $2.00 |
| Typical speed | 1.5 s / page | 3.0 s / page |
| Best for | Historical documents and manuscripts | Questions about document images |
| Inputs | images | images |
| Languages | Many scripts, incl. right-to-left | Multilingual |
| Max pages per request | 500 | 500 |
| Tables | No | No |
| Handwriting | ✓ Yes | No |
| Math | No | 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 Kraken if
- You are transcribing historical manuscripts or early prints, or working with Arabic, Hebrew and other RTL, BiDi or top-to-bottom scripts.
- You need ALTO, PageXML, abbyyXML or hOCR output for digital humanities tools, plus bounding boxes.
- You want to train layout analysis, reading order and character recognition on your own material, or start from community models on Zenodo.
Choose Pixtral if
- You already have an integration on the Mistral API and want to keep it.
- You need question answering over charts and documents, where Mistral reports 90.7 ANLS on DocVQA and 81.8 on ChartQA.
- You want one model that takes many images per request (128K token context) and returns plain text, Markdown or JSON, with Apache-2.0 open weights or a $0.15 per million token launch API price.
Key differences
| Kraken | Pixtral | |
|---|---|---|
| What it is | OCR system with trainable layout analysis and recognition, built for historical and non-Latin material | General vision language model (12B Mistral Nemo decoder plus 400M vision encoder) |
| How you run it | Python library, installed with pip on Linux or macOS (x64 or ARM, Python 3.10 to 3.13); no Windows support | Mistral API as pixtral-12b-2409, or self-hosted open weights |
| Licence and price | Apache-2.0; free, self-hosted (you pay for compute) | Apache-2.0; $0.15 per million input and output tokens at API launch, free to self-host |
| Output formats | ALTO, PageXML, hOCR, plain text, bounding boxes | Plain text, Markdown, JSON |
| Benchmarks | Not published | DocVQA 90.7 ANLS, ChartQA 81.8 (CoT), MMMU 52.5 (CoT), all reported by the authors |
| Handwriting and non-Latin scripts | Core focus, including RTL, BiDi and top-to-bottom support | Third-party guide reports an accuracy drop on handwriting, small fonts and non-Latin scripts |
| Known problems | Install can fail with a dependency conflict (issue #816); learned line order is lost in segmentation and text output (issue #820); usually needs training on your own material | Deprecated and no longer maintained; a launch-thread user said it does not always get OCR near correct; Q4 quantization reported to degrade small text OCR |
Sources
- Kraken GitHub README
- Zenodo OCR models
- Issue #816
- Mistral announcement
- Mistral Sept 2024 pricing
- Third-party deployment guide
What each one is
Kraken
OCR engine for historical documents and manuscripts, with trainable models for unusual scripts.
Kraken 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": "kraken/kraken","fallbacks": ["mistral/pixtral"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Kraken alternativesPixtral alternativesBest OCR models in 2026