OCRmyPDF vs Phi-4-multimodal
Pick OCRmyPDF to make scanned PDF archives searchable and PDF/A compliant, on CPU only. Pick Phi-4-multimodal to answer questions about English documents and to get Markdown or JSON back. The main split is the output: OCRmyPDF returns a searchable PDF or plain text, while Phi-4-multimodal reads the page as an image and returns text you can structure.
OCRmyPDFocrmypdf/ocrmypdf | Phi-4-multimodalmicrosoft/phi-4-multimodal | |
|---|---|---|
| Type | Open source | Vision model |
| Price | 1 credit / page | 2 credits / page |
| Per 1,000 pages | $1.00 | $2.00 |
| Typical speed | 1.0 s / page | 2.5 s / page |
| Best for | Searchable PDFs from scans | Small model for OCR and questions |
| Inputs | PDF, images | images |
| Languages | 100+ languages (via Tesseract) | English for vision |
| Max pages per request | 1,000 | 500 |
| Tables | No | ✓ Yes |
| Handwriting | No | No |
| Math | No | ✓ Yes |
| Bounding boxes | No | 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 OCRmyPDF if
- You need validated searchable PDF/A output for archiving.
- You have no GPU, since it runs on CPU only and spreads work across all cores.
- You have scans in many languages, since Tesseract language packs cover 100+ languages.
Choose Phi-4-multimodal if
- You want Markdown or JSON output, not only a text layer.
- You do English document Q&A, where Microsoft reports DocVQA 93.2 at 5.6B parameters.
- Your pipeline mixes audio and images, or you want an Azure AI Foundry option as well as self-hosting.
Key differences
| OCRmyPDF | Phi-4-multimodal | |
|---|---|---|
| Licence | MPL-2.0 | MIT |
| How you run it | Command-line tool that runs locally | Self-host the weights or use Azure AI Foundry |
| Hardware | CPU only, uses all cores | Flash attention, tested on A100, A6000 and H100. V100 and older need eager attention |
| Output formats | Searchable PDF (PDF/A by default) and plain text | Plain text, Markdown and JSON |
| Benchmarks | Not published | DocVQA 93.2, OCRBench 84.4, ChartQA 81.4, InfoVQA 72.7 (reported by the authors) |
| Languages | 100+ via Tesseract language packs | Vision supported for English only, and performance drops in other languages |
| Known failure | File size can double or quadruple with --deskew --clean-final (issue #948). --skip-text with default PDF/A output removes form fields (issue #1790). Hangs when forms share one Resources dictionary (issue #1763) | Model card warns of repetitive or inconsistent output in long chats |
Sources
- OCRmyPDF GitHub README
- OCRmyPDF docs
- Issue #948 file size
- Issue #1790 form fields
- Hugging Face model card
What each one is
OCRmyPDF
Command-line tool that adds a searchable text layer to scanned PDFs, using Tesseract.
OCRmyPDF detailsPhi-4-multimodal
Microsoft's 5.6B text, vision and audio model (February 2025). Scores 84.4 on OCRBench per its model card.
Phi-4-multimodal 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": "ocrmypdf/ocrmypdf","fallbacks": ["microsoft/phi-4-multimodal"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
OCRmyPDF alternativesPhi-4-multimodal alternativesBest OCR models in 2026