Amazon Textract vs docTR
Pick Amazon Textract if you are on AWS and process English forms, invoices or large scanned PDF batches, because it returns tables, forms and queries from a managed API. Pick docTR if you want a free, Apache-2.0 library you run yourself for word-level OCR with coordinates, and you can handle table or Markdown needs elsewhere. Textract is limited to six Latin-script languages; docTR gives you no table, formula or Markdown output at all.
Amazon Textractaws/textract | docTRmindee/doctr | |
|---|---|---|
| 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 | 0.9 s / page |
| Best for | Forms and key-value extraction | Word-level OCR with positions |
| Inputs | PDF, images | PDF, images |
| Languages | 6 languages | Mainly Latin scripts |
| Max pages per request | 3,000 | 500 |
| Tables | ✓ Yes | No |
| Handwriting | ✓ Yes | No |
| Math | No | No |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | ✓ Yes | ✓ Yes |
| 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
- It extracts tables ($0.015/page), forms ($0.05/page) and queries ($0.015/page), which docTR does not return.
- The async API takes PDFs up to 500 MB and 3,000 pages, so large scanned batches need no infrastructure from you.
- It has a published table accuracy score of 80.9% on RD-TableBench, and returns word and line geometry with confidence scores.
Choose docTR if
- It is Apache-2.0 and free to self-host, so you pay only for compute and no per-page fee applies.
- You choose the detection and recognition architectures, with straight or rotated boxes, and can train your own models in Python.
- It exports a nested page, block, line and word structure with coordinates as JSON, which suits invoices and receipts feeding custom extraction.
Key differences
| Amazon Textract | docTR | |
|---|---|---|
| Licence | Proprietary | Apache-2.0 |
| How you run it | Hosted AWS API | Python library on PyTorch, with Docker images; CPU or NVIDIA GPU (GPU images need CUDA 12.2 or newer on the host) |
| Price | US West (Oregon), first 1M pages/month: text detection $0.0015/page, Tables $0.015, Forms $0.05, Queries $0.015 | Free, self-hosted (you pay for compute) |
| Tables and forms | Tables, forms, queries and layout supported | No table, formula or Markdown output; words and lines only |
| Languages | Text detection: English, French, German, Italian, Portuguese, Spanish. Handwriting and Queries: English only. Vertical Japanese and Chinese not supported | Not published |
| Benchmark | 80.9% on RD-TableBench table accuracy (independent) | Not published |
| Known failure | Sync API takes one page only | Cropping images gives poorer results (issue #2007); split-crop merging can misjoin readings across repeated characters (issue #2229); ONNX export of MASTER truncates decoding steps (issue #2224) |
Sources
- Textract pricing
- Textract quotas
- Textract hard limits
- Reducto RD-TableBench
- docTR GitHub README
- Issue #2007
What each one is
Amazon Textract
AWS OCR with form key-value pairs, tables and per-word confidence scores.
Amazon Textract detailsdocTR
Open-source OCR library from Mindee with separate text detection and recognition models.
docTR 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": ["mindee/doctr"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Amazon Textract alternativesdocTR alternativesBest OCR models in 2026