Marker vs Nougat
Pick Marker for self-hosted PDF to Markdown across mixed documents, RAG chunking, or when you may want a hosted API. Pick Nougat only for English, math-heavy arXiv-style papers where you need LaTeX output and the use is non-commercial. The main split is scope and licence: Marker has several output formats and a startup-friendly licence, while Nougat is limited to scientific papers and its weights bar commercial use.
Markerdatalab/marker | Nougatmeta/nougat | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 14 credits / page | 2 credits / page |
| Per 1,000 pages | $14.00 | $2.00 |
| Typical speed | 2.5 s / page | 3.0 s / page |
| Best for | Books and research papers | arXiv-style papers |
| Inputs | PDF, images, DOCX | PDF, images |
| Languages | 90+ languages | Mainly English |
| Max pages per request | 1,000 | 200 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| Math | ✓ Yes | ✓ 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 Marker if
- You need output beyond Markdown and LaTeX: HTML, JSON block trees, RAG chunks or images.
- You have no GPU: fast mode runs on CPU with about 18 MB of ONNX weights at 23.6 pages per second, and it also runs on Apple Silicon.
- You want a hosted option: Datalab's API includes about 5,000 free pages per month.
Choose Nougat if
- Your input is English arXiv or PubMed Central style papers, which is what it was trained on.
- You want equations and tables as LaTeX in Mathpix-compatible .mmd Markdown.
- You want a small package with small and base checkpoints and a simple CLI and API, and your use is non-commercial.
Key differences
| Marker | Nougat | |
|---|---|---|
| Licence | Code Apache-2.0; weights under a modified AI Pubs Open Rail-M licence (free for research, personal use and startups under $5M) | Code MIT; weights CC-BY-NC, so no commercial use |
| How you run it | Self-host on GPU, CPU or Apple Silicon (MPS), or use Datalab's hosted API | Self-hosted Python package only |
| Price | Free self-hosted within the weight licence; hosted API has about 5,000 free pages per month, then paid plans | Free for non-commercial use, self-hosted |
| Output formats | Markdown, HTML, JSON, chunks, LaTeX, images | Markdown (.mmd) and LaTeX |
| Benchmark | 83.7 on olmOCR-bench in vllm mode, 66.6 in CPU fast mode (Datalab's own numbers) | Not published |
| Languages | Not published | Works best on English; README says Chinese, Russian and Japanese do not work |
| Known failure | CPU fast mode scores lower (66.6 vs 83.7 in vllm mode) | Repetition loops on 1.5% of test pages, more on out-of-domain documents; repetition detection can drop pages as MISSING_PAGE |
Sources
What each one is
Marker
Hosted version of the open-source Marker and Surya models. Strong on books and papers.
Marker 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": ["meta/nougat"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Marker vs DonutMarker vs MonkeyOCRAmazon Textract vs NougatReducto Parse vs NougatLlamaParse vs MarkerMarker vs Nanonets-OCRJina OCR vs NougatMistral OCR vs Nougat
Marker alternativesNougat alternativesBest OCR models in 2026