Florence-2 vs Unstructured
Pick Florence-2 for scene text and short English labels where you need bounding boxes from a 0.23B or 0.77B model, or when the same model must also do detection. Pick Unstructured for ingesting mixed files such as PDFs, Word, PowerPoint, emails and HTML into typed elements for RAG. The main split is scope: Florence-2 returns only text and boxes, while Unstructured partitions whole documents.
Florence-2microsoft/florence-2 | Unstructuredunstructured/unstructured | |
|---|---|---|
| Type | Open source | OCR API |
| Price | 1 credit / page | 20 credits / page |
| Per 1,000 pages | $1.00 | $20.00 |
| Typical speed | 0.8 s / page | 3.0 s / page |
| Best for | Lightweight OCR with boxes | Mixed file types into RAG |
| Inputs | images | PDF, images, DOCX |
| Languages | English | Multilingual |
| Max pages per request | 500 | 1,000 |
| Tables | No | ✓ Yes |
| Handwriting | No | No |
| Math | No | No |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | No | No |
| Forms | No | No |
| Multilingual | No | ✓ 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 Florence-2 if
- It is small (0.23B base, 0.77B large) and MIT licensed, so it fits lightweight edge setups where you pay only for compute.
- OCR_WITH_REGION returns quadrilateral boxes per text line, which suits scene text and short labels.
- The same model also does captioning, detection and grounding, so one model can serve a pipeline that needs more than OCR.
Choose Unstructured if
- It handles PDFs, HTML, Word, PowerPoint, emails and images in one library, so mixed-format ingestion needs one tool.
- The hosted API gives 10,000 pages free, then $0.015 per page, so you can try it before paying.
- Its Auto, Fast, High-Res and VLM strategies let you trade speed for quality, and Fast suits born-digital PDFs.
Key differences
| Florence-2 | Unstructured | |
|---|---|---|
| Licence | MIT | Apache-2.0 for the open-source library; proprietary for the hosted platform |
| How you run it | Self-hosted via Transformers | Self-hosted library or hosted API |
| Price | Free, you pay for compute | Library free; hosted is 10,000 pages free, then $0.015 per page |
| Output formats | Plain text and bounding boxes; no Markdown, tables or LaTeX | JSON, HTML, plain text and bounding boxes |
| Input types | Images; also used for captioning, detection and grounding | PDFs, HTML, Word, PowerPoint, emails and images |
| Tables | No table structure output | Open-source table cell content scores 0.426 versus 0.866 hosted; hi_res tables can cut off edges and drop first letters; table OCR ignores language settings |
| Known failure | Spelling errors and missing spaces on cursive handwriting and French text; Chinese output mostly wrong because the vocabulary lacks CJK characters | Extra Tesseract pass on tables adds processing time; the library sends usage telemetry by default |
Sources
- Florence-2-large model card
- HF blog OCR test
- HF discussion Chinese very bad
- Pricing
- GitHub README
- Issue #3262
What each one is
Florence-2
Small Microsoft vision model (0.23B and 0.77B) with OCR and OCR-with-regions tasks that return text with boxes.
Florence-2 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": "microsoft/florence-2","fallbacks": ["unstructured/unstructured"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Florence-2 alternativesUnstructured alternativesBest OCR models in 2026