Marker vs Kraken
Pick Marker to turn modern PDFs, books and papers into Markdown, HTML or JSON for RAG. Pick Kraken for historical manuscripts, early prints and non-Latin or right-to-left scripts, where you can train on your own material. Marker is closer to plug and play, while Kraken is a trainable toolkit for research projects.
Markerdatalab/marker | Krakenkraken/kraken | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 14 credits / page | 1 credit / page |
| Per 1,000 pages | $14.00 | $1.00 |
| Typical speed | 2.5 s / page | 1.5 s / page |
| Best for | Books and research papers | Historical documents and manuscripts |
| Inputs | PDF, images, DOCX | images |
| Languages | 90+ languages | Many scripts, incl. right-to-left |
| Max pages per request | 1,000 | 500 |
| Tables | ✓ Yes | No |
| Handwriting | No | ✓ Yes |
| Math | ✓ Yes | No |
| Bounding boxes | ✓ Yes | ✓ Yes |
| 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 Marker if
- You want PDF to Markdown, with HTML, JSON block trees, RAG chunks, LaTeX and images as other output options.
- You work with books and papers with equations, and Datalab reports 83.7 on olmOCR-bench in vllm mode.
- You want a choice of local use on GPU, CPU or Apple Silicon, or Datalab's hosted API with about 5,000 free pages per month.
Choose Kraken if
- You transcribe historical manuscripts or early prints, or scripts such as Arabic and Hebrew, with right-to-left, BiDi and top-to-bottom support.
- You need ALTO, PageXML, abbyyXML or hOCR output for digital humanities tools, and want to train layout analysis, reading order and character recognition yourself.
- You want a fully Apache-2.0 library and can use the community models on Zenodo as a starting point.
Key differences
| Marker | Kraken | |
|---|---|---|
| Main focus | PDF to Markdown conversion, good for books and papers with equations, and RAG chunking | Historical documents and non-Latin scripts, with trainable layout analysis and recognition |
| Licence | Code Apache-2.0. Weights under a modified AI Pubs Open Rail-M licence: free for research, personal use and startups under $5M. | Apache-2.0 |
| How you run it | Locally on GPU, CPU or Apple Silicon, or through Datalab's hosted API. Fast mode runs on CPU with about 18 MB of ONNX weights. | Python library installed with pip on Linux or macOS (x64 or ARM, Python 3.10 to 3.13). No Windows support. |
| Price | Free self-hosted within the weight licence. Hosted API has about 5,000 free pages per month, then paid plans. | Free, self-hosted. You pay for compute. |
| Output formats | Markdown, HTML, JSON, chunks, LaTeX, images | ALTO, PageXML, hOCR, plain text, bounding boxes |
| Benchmarks | olmOCR-bench 83.7 and OmniParseBench 85.9 in vllm mode. CPU fast mode scores 66.6 on olmOCR-bench. All figures are Datalab's own. | Not published |
| Known problems | Weights licence blocks free commercial use for companies above $5M. CPU fast mode is less accurate. | 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. |
Sources
What each one is
Marker
Hosted version of the open-source Marker and Surya models. Strong on books and papers.
Marker detailsKraken
OCR engine for historical documents and manuscripts, with trainable models for unusual scripts.
Kraken detailsTry both on your own file
Run the file once with each model and compare the output.
PDF, PNG, JPG, WebP, HEIC. Up to 20 MB each, 10 files.No file?
Use one, fall back to the other
{"model": "datalab/marker","fallbacks": ["kraken/kraken"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Marker vs TrOCRMarker vs docTRAmazon Textract vs KrakenReducto Parse vs KrakenMarker vs MiniCPM-V 4.5Marker vs RapidOCRdocTR vs KrakenDeepSeek-OCR vs Kraken
Marker alternativesKraken alternativesBest OCR models in 2026