Mistral OCR vs TrOCR
Pick Mistral OCR when you need whole pages or PDFs turned into Markdown, including tables, in batch. Pick TrOCR when you already have cropped single text-line images, mainly English handwriting or receipt lines, and want a free, self-hosted recognizer you can fine-tune. The main split is scope: Mistral OCR reads full documents through an API, TrOCR only reads one line at a time.
Mistral OCRmistral/mistral-ocr | TrOCRmicrosoft/trocr | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 25 credits / page | 1 credit / page |
| Per 1,000 pages | $25.00 | $1.00 |
| Typical speed | 0.9 s / page | 1.2 s / page |
| Best for | High-volume PDF to Markdown | Handwritten text lines |
| Inputs | PDF, images | images |
| Languages | 100+ languages | English |
| Max pages per request | 1,000 | 100 |
| Tables | ✓ Yes | No |
| Handwriting | No | ✓ Yes |
| Math | ✓ Yes | No |
| 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 Mistral OCR if
- It takes full pages and PDFs and returns Markdown, HTML, JSON and extracted images, with tables rebuilt as HTML with rowspan and colspan in OCR 3.
- It costs $2 per 1,000 pages, or $1 per 1,000 through the Batch API, which suits high-volume PDF to Markdown for RAG and batches of scanned business documents.
- You call a hosted API, so you run no GPU and need no line detector in front of it.
Choose TrOCR if
- It scores 2.89 character error rate on the IAM handwriting set with the large model, per the paper, so it fits English handwritten lines.
- It is MIT licensed and free to self-host, and you can fine-tune it on your own line images.
- It has separate checkpoints for handwritten (IAM), printed (SROIE) and scene text, and loads in a few lines with Hugging Face transformers.
Key differences
| Mistral OCR | TrOCR | |
|---|---|---|
| Input scope | Full pages and PDFs | Single text-line images only; you need a separate line detector |
| Deployment | Hosted API (mistral-ocr-2512, OCR 3) | Library, run locally through Hugging Face transformers |
| Licence | Proprietary | MIT (microsoft/unilm repository) |
| Price | $2 per 1,000 pages; $3 with structured annotations; $1 via Batch API | Free, self-hosted; you pay for compute |
| Output | Markdown, HTML, JSON, images | Plain text only, no layout or table output |
| Languages | Not published | Released checkpoints are English-focused; other languages need fine-tuning |
| Known failure | Hallucinated text on unreadable handwriting; changed checkbox states and misread characters on a dense financial table page | Up to 558M parameters for one text line, slow versus CRNN recognizers on CPU |
Sources
- Mistral OCR 3 announcement
- VentureBeat on OCR 3 pricing
- Reducto test of Mistral OCR
- TrOCR paper (arXiv 2109.10282)
- trocr-large-handwritten model card
- Transformers TrOCR docs
What each one is
Mistral OCR
Fast document OCR that returns clean Markdown with tables and embedded images.
Mistral OCR detailsTrOCR
Transformer text recognition model. Reads single text lines, including handwriting.
TrOCR 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": "mistral/mistral-ocr","fallbacks": ["microsoft/trocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Mistral OCR alternativesTrOCR alternativesBest OCR models in 2026