Kraken vs olmOCR
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 olmOCR for bulk PDF to Markdown conversion of academic papers and mixed printed and handwritten pages, if you have an NVIDIA GPU with 12 GB or more of VRAM. The main split is document type: Kraken targets historical and non-Latin material, olmOCR targets modern PDFs for LLM training or RAG.
Krakenkraken/kraken | olmOCRallenai/olmocr | |
|---|---|---|
| Type | Open source | Open source |
| Price | 1 credit / page | 2 credits / page |
| Per 1,000 pages | $1.00 | $2.00 |
| Typical speed | 1.5 s / page | 1.5 s / page |
| Best for | Historical documents and manuscripts | Bulk PDF to text for datasets |
| Inputs | images | PDF, images |
| Languages | Many scripts, incl. right-to-left | Mainly English |
| Max pages per request | 500 | 1,000 |
| Tables | No | ✓ Yes |
| Handwriting | ✓ Yes | ✓ Yes |
| Math | No | ✓ Yes |
| Bounding boxes | ✓ Yes | No |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | ✓ Yes | No |
| 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 work with historical manuscripts and early prints, or with Arabic, Hebrew and other right-to-left, BiDi or top-to-bottom scripts.
- You need ALTO, PageXML, abbyyXML or hOCR output for digital humanities tools, and want to train layout analysis, reading order and recognition yourself.
- You have no GPU and run Linux or macOS, and you can start from the community-trained models on Zenodo.
Choose olmOCR if
- You need clean Markdown from large volumes of PDFs for LLM training or RAG, at under $200 per million pages self-hosted, per Ai2.
- Your documents are academic papers with equations, tables, or pages that mix print and handwriting, and you want headers and footers stripped in reading order.
- You want a 7B model with open weights, training data and code, or a hosted option such as DeepInfra at $0.09 input and $0.19 output per million tokens.
Key differences
| Kraken | olmOCR | |
|---|---|---|
| Best document types | Historical manuscripts, early prints, non-Latin and right-to-left scripts | PDFs and images to Markdown: academic papers, tables, equations, mixed print and handwriting |
| 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 | Self-hosted with vLLM on an NVIDIA GPU with at least 12 GB VRAM and 30 GB free disk, or via third-party hosted endpoints |
| Licence | Apache-2.0 | Apache-2.0 (toolkit); weights, training data and code are released |
| Price | Free, self-hosted; you pay for compute | Self-hosted under $200 per million pages, per Ai2; DeepInfra olmOCR-2-7B $0.09 input / $0.19 output per million tokens |
| Output formats | ALTO, PageXML, hOCR, plain text, bounding boxes | Markdown, JSON |
| Benchmark | Not published | 82.4 on olmOCR-Bench (v0.4.0), above Marker (76.1), MinerU 2.5.4 (75.2) and Mistral OCR (72.0) |
| Known problems | Install dependency conflict (issue #816); learned line order lost in segmentation and text output (issue #820); usually needs training on your own material | Out of memory on a 12 GB GPU (issue #424); gets stuck on the last few pages (issue #433); unreliable image bounding box coordinates (issue #437) |
Sources
What each one is
Kraken
OCR engine for historical documents and manuscripts, with trainable models for unusual scripts.
Kraken detailsolmOCR
Open-source vision model from Ai2 (often searched as olm ocr) that turns PDFs into clean, reading-order text. Latest release: olmOCR 2.
olmOCR 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": ["allenai/olmocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Kraken alternativesolmOCR alternativesBest OCR models in 2026