Google Document AI vs Unlimited OCR
Pick Google Document AI if you run on Google Cloud and need a managed API for bulk OCR at $1.50 per 1,000 pages, or Layout Parser chunking for RAG. Pick Unlimited OCR if you want an MIT-licensed model you host yourself to parse long reports and books, 40+ pages in a single pass, into Markdown with bounding boxes. The main split is hosted per-page pricing with request limits versus free weights that need an NVIDIA GPU and can loop or hallucinate.
Google Document AIgoogle/document-ai | Unlimited OCRbaidu/unlimited-ocr | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 10 credits / page | 2 credits / page |
| Per 1,000 pages | $10.00 | $2.00 |
| Typical speed | 1.4 s / page | 1.2 s / page |
| Best for | Multilingual scans | Books and very long documents |
| Inputs | PDF, images | PDF, images |
| Languages | 200+ languages | 30+ languages |
| Max pages per request | 500 | 1,000 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | No | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | ✓ Yes | No |
| Forms | ✓ Yes | 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 Document AI if
- Enterprise Document OCR costs $1.50 per 1,000 pages, drops to $0.60 above 5M pages a month, and the first 1,000 pages are free.
- Layout Parser returns chunks for RAG at $10 per 1,000 pages, so you do not build the chunking step yourself.
- It is a hosted API, so you need no GPU, and batch requests take up to 500 pages.
Choose Unlimited OCR if
- The MIT licence allows commercial use with no copyleft, and the software is free if you self-host (you pay only for compute).
- It parses 40+ pages in one forward pass with roughly constant memory, while Document AI online requests stop at 15 pages (30 in imageless mode).
- It outputs Markdown with block types and bounding boxes, which suits layout-aware extraction and bulk PDF to text for RAG.
Key differences
| Google Document AI | Unlimited OCR | |
|---|---|---|
| How you run it | Hosted API on Google Cloud | Self-host with vLLM, SGLang or transformers; Baidu Cloud also offers a hosted endpoint |
| Licence | Proprietary | MIT |
| Price | $1.50 per 1,000 pages for Enterprise OCR, $0.60 above 5M a month; Layout Parser $10; Form Parser $30 | Free to self-host, you pay for compute; hosted endpoint pricing not published |
| Pages per request | Online 15 pages (30 imageless); batch up to 500; Form Parser batch 100 | 40+ pages in a single pass, context capped at 32,768 tokens, multi-page input only in base mode |
| Output formats | JSON, plain text, bounding boxes, chunks | Markdown, plain text, bounding boxes |
| Table accuracy | 64.6% on RD-TableBench, behind Azure and Textract | Not published |
| Known failures | Weak on complex tables; online limits make long PDFs awkward in one synchronous call | Text looping (issue #55), hallucinated text (issue #58), bbox coordinate loops under SGLang on GB10 GPUs (issue #114), mostly loses bold and italic styling, 12 GB GPUs run out of memory with version 2 (issue #116) |
Sources
- Document AI pricing
- Document AI limits
- Reducto RD-TableBench
- Hugging Face model card
- Issue #55 repetition
- Issue #116 OOM on 12 GB
What each one is
Google Document AI
Google Cloud OCR with layout parsing and specialised processors.
Google Document AI detailsUnlimited OCR
Open-source model from Baidu that reads hundreds of pages in one pass with flat memory use.
Unlimited OCR 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/document-ai","fallbacks": ["baidu/unlimited-ocr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Google Document AI alternativesUnlimited OCR alternativesBest OCR models in 2026