Marker vs Unstructured
Pick Marker if you want to self-host PDF to Markdown conversion for books, papers with equations, or RAG chunking, and your company is under the $5M weight-licence threshold or will pay for a licence. Pick Unstructured if you need to ingest mixed file types (PDF, HTML, Word, PowerPoint, email, images) through one library with an Apache-2.0 licence. The main split is scope: Marker is a focused PDF converter, Unstructured is a multi-format ingestion pipeline.
Markerdatalab/marker | Unstructuredunstructured/unstructured | |
|---|---|---|
| Type | OCR API | OCR API |
| Price | 14 credits / page | 20 credits / page |
| Per 1,000 pages | $14.00 | $20.00 |
| Typical speed | 2.5 s / page | 3.0 s / page |
| Best for | Books and research papers | Mixed file types into RAG |
| Inputs | PDF, images, DOCX | PDF, images, DOCX |
| Languages | 90+ languages | Multilingual |
| Max pages per request | 1,000 | 1,000 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | No | No |
| 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 Markdown, HTML, JSON block trees, LaTeX or RAG chunks as output from PDFs.
- You need to run on CPU, Apple Silicon or GPU, including a fast CPU mode with about 18 MB of ONNX weights at 23.6 pages per second.
- Your documents are books or papers with equations, and Datalab reports 83.7 on olmOCR-bench in vllm mode.
Choose Unstructured if
- You ingest many file types, including PDFs, HTML, Word, PowerPoint, emails and images, with one library.
- You want a fully Apache-2.0 open-source library with many connectors, plus 10,000 free hosted pages to start.
- Your PDFs are born-digital and the fast strategy gives you the speed you need, with Auto, Fast, High-Res and VLM strategies to trade speed for quality.
Key differences
| Marker | Unstructured | |
|---|---|---|
| Licence | Code Apache-2.0; model weights under a modified AI Pubs Open Rail-M licence, free for research, personal use and startups under $5M | Apache-2.0 for the open-source library; proprietary for the hosted platform |
| Price | Free self-hosted within the weight licence (you pay for compute); hosted API has about 5,000 free pages per month, paid plans on Datalab's pricing page | Library is free; hosted gives the first 10,000 pages free, then $0.015 per page pay-as-you-go; Business is custom |
| Input scope | PDF to Markdown converter | PDFs, HTML, Word, PowerPoint, emails and images |
| Output formats | Markdown, HTML, JSON, chunks, LaTeX, images | JSON, HTML, plain text, bounding boxes |
| Benchmark | 83.7 on olmOCR-bench (vllm mode) and 85.9 on OmniParseBench, all reported by Datalab; CPU fast mode drops to 66.6 on olmOCR-bench | Table cell content 0.426 for the open-source library versus 0.866 hosted, reported by the vendor |
| Tables and languages | Not published | hi_res tables can cut off at the edges and drop first letters; table OCR ignores language settings; an extra Tesseract pass on tables adds processing time |
| Telemetry and commercial limits | Weight licence blocks free commercial use for companies above $5M without a paid licence | Library sends usage telemetry by default |
Sources
- Marker GitHub README
- Datalab pricing
- Unstructured pricing
- Unstructured GitHub README
- Unstructured issue #3262
- Unstructured issue #1694
What each one is
Marker
Hosted version of the open-source Marker and Surya models. Strong on books and papers.
Marker detailsUnstructured
Document ETL for RAG. Partitions many file types into typed elements; open-source library too.
Unstructured 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": "datalab/marker","fallbacks": ["unstructured/unstructured"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Marker alternativesUnstructured alternativesBest OCR models in 2026