Mistral OCR vs Surya
Pick Mistral OCR if you want a hosted API for high-volume PDF to Markdown at $2 per 1,000 pages ($1 via Batch API), with no servers to run. Pick Surya if documents must stay on your own hardware, you need layout labels, reading order and bounding boxes, or your revenue and funding are under $5M. Avoid both for messy handwriting.
Mistral OCRmistral/mistral-ocr | Suryadatalab/surya | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 25 credits / page | 2 credits / page |
| Per 1,000 pages | $25.00 | $2.00 |
| Typical speed | 0.9 s / page | 1.5 s / page |
| Best for | High-volume PDF to Markdown | Layout and line-level OCR |
| Inputs | PDF, images | PDF, images |
| Languages | 100+ languages | 90+ languages |
| Max pages per request | 1,000 | 500 |
| 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 Mistral OCR if
- You want a hosted API with no GPU or inference server to run, priced at $2 per 1,000 pages and $1 per 1,000 through the Batch API.
- Your documents have tables, since OCR 3 rebuilds them as HTML with rowspan and colspan.
- You process scanned business documents or PDFs for RAG in large batches and want Markdown, HTML or JSON back.
Choose Surya if
- You need to run locally: on an NVIDIA GPU via vLLM, or on CPU and Apple Silicon via llama.cpp with about 4-6 GB peak RAM on an M4 Max.
- You need layout labels, reading order, and per-block bounding boxes across 90+ languages from one 689M-parameter model.
- You want no per-page fee. Code is Apache-2.0, and the weights are free for research, personal use and companies under $5M funding or revenue.
Key differences
| Mistral OCR | Surya | |
|---|---|---|
| How you run it | Hosted Mistral API, pinned version mistral-ocr-2512 (OCR 3) | Self-hosted; a lightweight client talks to a vLLM or llama.cpp server you run |
| Licence | Proprietary | Code Apache-2.0; weights under a modified AI Pubs Open Rail-M, with a commercial licence needed above $5M funding or revenue |
| Price | $2 per 1,000 pages; $3 with structured annotations; $1 via Batch API | Free software; you pay for compute |
| Output formats | Markdown, HTML, JSON, images | JSON, HTML, LaTeX, bounding boxes, plain text |
| Benchmark | OmniDocBench v1.5: 78.83 for the first version (OCR 2503); v1.6: 85.66 (independent) | olmOCR-bench 83.7% and OmniParseBench 85.9 for Surya 2.5 (reported by the authors) |
| Tables | HTML tables with rowspan and colspan | Surya 2.5 table output lacks row and column geometry; multilingual tables score 58.1 |
| Known failure | Hallucinated text on unreadable handwriting; changed checkbox states and misread characters on a dense financial table page | Handwriting, complex tables and math are better sent to the hosted Datalab API; not for photos or natural scenes |
Sources
- Mistral OCR 3 announcement
- VentureBeat on OCR 3 pricing
- OmniDocBench v1.6 leaderboard
- Reducto test of Mistral OCR
- Surya GitHub README
- Datalab pricing and license FAQ
What each one is
Mistral OCR
Fast document OCR that returns clean Markdown with tables and embedded images.
Mistral OCR detailsSurya
Open-source OCR toolkit with line detection, layout analysis, reading order and tables.
Surya 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": "mistral/mistral-ocr","fallbacks": ["datalab/surya"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Mistral OCR vs dots.ocrMistral OCR vs ChandraGemini Flash vs SuryaSurya vs ERNIE 4.5 VLMistral OCR vs Florence-2MinerU vs SuryaSurya vs MiniCPM-V 4.5Chandra vs Surya
Mistral OCR alternativesSurya alternativesBest OCR models in 2026