Amazon Textract vs Granite Docling
Pick Amazon Textract for English forms, invoices and large scanned PDF batches on AWS, where a managed API with per-page pricing and 3,000-page async jobs matters. Pick Granite Docling for self-hosted conversion of English PDFs with tables, code and equations into Markdown or HTML at the cost of compute only. The main split is a hosted proprietary API versus a free Apache-2.0 258M model you run yourself.
Amazon Textractaws/textract | Granite Doclingibm/granite-docling | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 25 credits / page | 1 credit / page |
| Per 1,000 pages | $25.00 | $1.00 |
| Typical speed | 1.2 s / page | 1.5 s / page |
| Best for | Forms and key-value extraction | Small-model document conversion |
| Inputs | PDF, images | PDF, images |
| Languages | 6 languages | English, experimental CJK and Arabic |
| Max pages per request | 3,000 | 500 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | No |
| Math | No | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | ✓ Yes | No |
| Forms | ✓ Yes | No |
| Multilingual | No | 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 Amazon Textract if
- Basic text detection costs $1.50 per 1,000 pages, with Tables at $0.015 per page and Forms at $0.05 per page in US West (Oregon) for the first 1M pages a month.
- The async API takes PDFs up to 500 MB and 3,000 pages, so large scanned batches need no splitting.
- It returns word and line geometry with confidence scores, and handles printed and handwritten English text, forms and queries.
Choose Granite Docling if
- It is Apache-2.0 licensed and free to self-host, so you pay only for compute and can run it on CPU or a small GPU.
- It outputs DocTags, Markdown, HTML, LaTeX and bounding boxes, so you get Markdown for RAG without post-processing.
- The model card reports table structure TEDS of 0.97 on FinTabNet, code recognition F1 of 0.988 and equation F1 of 0.968.
Key differences
| Amazon Textract | Granite Docling | |
|---|---|---|
| How you run it | Hosted AWS API | Self-hosted, typically inside Docling |
| Licence | Proprietary | Apache-2.0 |
| Price | $0.0015/page for text detection, $0.015/page for Tables, $0.05/page for Forms (US West, first 1M pages/month) | Free, you pay for compute |
| Output formats | JSON, plain text, bounding boxes | DocTags, Markdown, HTML, LaTeX, bounding boxes |
| Table benchmark | 80.9% on RD-TableBench, measured independently | TEDS 0.97 on FinTabNet at 150dpi, reported by the model authors (different benchmark, not directly comparable) |
| Languages | Text detection: English, French, German, Italian, Portuguese, Spanish; handwriting and Queries English only | English primary; Japanese, Arabic and Chinese marked experimental |
| Known failure | Vertical Japanese and Chinese text not supported; sync API takes one page | Can get stuck in repetitive loops on page headings and invoices; may output only exclamation marks on GPUs without bfloat16 such as T4 unless run in float32 |
Sources
- Textract pricing
- Textract quotas
- Reducto RD-TableBench
- Granite-Docling model card
- HF discussion #24 repetition
What each one is
Amazon Textract
AWS OCR with form key-value pairs, tables and per-word confidence scores.
Amazon Textract detailsGranite Docling
Compact IBM vision model that converts pages to DocTags for the Docling pipeline.
Granite Docling 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": "aws/textract","fallbacks": ["ibm/granite-docling"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Amazon Textract alternativesGranite Docling alternativesBest OCR models in 2026