Google Cloud Vision OCR vs Unstructured
Pick Google Cloud Vision OCR for high-volume plain text extraction from photos, signs, labels and handwritten notes, where you want word-level bounding polygons and a low per-page price. Pick Unstructured to ingest mixed files such as PDFs, Word, PowerPoint and email into typed elements for RAG, or if you need to self-host. Cloud Vision has no table or layout structure, so tables are the main reason to look elsewhere.
Google Cloud Vision OCRgoogle/cloud-vision | Unstructuredunstructured/unstructured | |
|---|---|---|
| Type | OCR API | OCR API |
| Price | 2 credits / page | 20 credits / page |
| Per 1,000 pages | $2.00 | $20.00 |
| Typical speed | 1.0 s / page | 3.0 s / page |
| Best for | Photos and simple scans | Mixed file types into RAG |
| Inputs | PDF, images | PDF, images, DOCX |
| Languages | 100+ languages | Multilingual |
| Max pages per request | 2,000 | 1,000 |
| Tables | No | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | No | No |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | ✓ Yes | 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 Google Cloud Vision OCR if
- It returns a page, block, paragraph, word and symbol hierarchy with bounding polygons, so you get fine-grained positions for text.
- It supports handwriting through DOCUMENT_TEXT_DETECTION and language hints, which suits handwritten notes.
- Pricing falls to $0.60 per 1,000 units above 5M units, and each PDF page counts as one unit, so it suits high volume.
Choose Unstructured if
- It handles PDFs, HTML, Word, PowerPoint, emails and images in one library, so one pipeline covers mixed file types.
- You can self-host the Apache-2.0 open-source library for free, or call the hosted API, which gives the first 10,000 pages free.
- Its strategies Auto, Fast, High-Res and VLM let you trade speed for quality, and Fast works for born-digital PDFs.
Key differences
| Google Cloud Vision OCR | Unstructured | |
|---|---|---|
| How you run it | Hosted API only, called over REST or client libraries on Google Cloud. No on-premise deployment. | Self-host the Python library or use the hosted API. |
| Licence | Proprietary | Apache-2.0 for the open-source library; proprietary for the hosted platform |
| Price | First 1,000 units per month free per feature, then $1.50 per 1,000 units up to 5M, then $0.60 per 1,000 above. Billing must be enabled even for the free units. | Hosted: first 10,000 pages free, then $0.015 per page. Open-source library is free. Business is custom. |
| Output formats | JSON, plain text, bounding boxes. No Markdown output, so you rebuild reading order yourself. | JSON, HTML, plain text, bounding boxes. |
| Tables and layout | No table or layout structure. Google points to Document AI for forms and tables. | Partitions files into typed elements. Vendor's own benchmark for table cell content: 0.426 open-source vs 0.866 hosted. |
| PDF handling | PDF and TIFF require async processing via Cloud Storage. Async batch accepts up to 2,000 image files per request. | Accepts PDFs directly alongside HTML, Word, PowerPoint, emails and images. |
| Known failure | Not published | hi_res tables cut off at edges and drop first letters (issue #3262). Table OCR ignores language settings (issue #1694). The library sends usage telemetry by default. |
Sources
What each one is
Google Cloud Vision OCR
Google Cloud's image OCR with text and document text detection, word boxes and confidence.
Google Cloud Vision OCR 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": "google/cloud-vision","fallbacks": ["unstructured/unstructured"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Google Cloud Vision OCR alternativesUnstructured alternativesBest OCR models in 2026