Amazon Textract vs Qwen-VL
Pick Amazon Textract for English forms, invoices and large scanned PDF batches on AWS, where you want a managed API with per-page pricing and word-level confidence scores. Pick Qwen-VL for Chinese, Japanese and other non-Latin text, for photos and screenshots, or when you want one open model that also answers questions about the page. Qwen-VL needs loop checks on low-quality scans; Textract does not support those languages.
Amazon Textractaws/textract | Qwen-VLqwen/qwen-vl | |
|---|---|---|
| Type | OCR API | Open source |
| Price | 25 credits / page | 2 credits / page |
| Per 1,000 pages | $25.00 | $2.00 |
| Typical speed | 1.2 s / page | 2.0 s / page |
| Best for | Forms and key-value extraction | Flexible extraction with prompts |
| Inputs | PDF, images | PDF, images |
| Languages | 6 languages | 30+ languages |
| Max pages per request | 3,000 | 200 |
| Tables | ✓ Yes | ✓ Yes |
| Handwriting | ✓ Yes | ✓ Yes |
| Math | No | ✓ Yes |
| Bounding boxes | ✓ Yes | ✓ Yes |
| Confidence | ✓ Yes | No |
| Forms | ✓ Yes | 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 Amazon Textract if
- It is a hosted API with published per-page prices: $0.0015 for basic text detection, $0.015 for tables and $0.05 for forms in US West (Oregon), first 1M pages a month.
- The async API takes PDFs up to 500 MB and 3,000 pages, which suits large scanned batches.
- It returns word and line geometry with confidence scores, and it scored 80.9% on RD-TableBench table accuracy.
Choose Qwen-VL if
- It runs OCR in 32 languages, while Textract text detection covers only English, French, German, Italian, Portuguese and Spanish.
- It is Apache-2.0 with open weights from 2B to 235B-A22B MoE, so you can self-host and pay only for compute.
- It outputs Markdown, HTML, JSON and bounding boxes, and handles photos and screenshots with text as well as OCR plus reasoning or VQA.
Key differences
| Amazon Textract | Qwen-VL | |
|---|---|---|
| How you run it | Hosted AWS API only | Self-hosted open weights, or a hosted API on Alibaba Cloud |
| Licence | Proprietary | Apache-2.0 (Qwen3-VL repository) |
| Price | $0.0015/page for text detection, $0.015 tables, $0.05 forms (US West Oregon, first 1M pages/month) | Free self-hosted, you pay for compute. Hosted pricing Not published |
| Languages | English, French, German, Italian, Portuguese, Spanish for text detection. Handwriting and Queries English only. Vertical Japanese and Chinese not supported. | OCR in 32 languages |
| Output formats | JSON, plain text, bounding boxes | Plain text, HTML, Markdown, JSON, bounding boxes |
| Benchmark | 80.9% on RD-TableBench table accuracy | Qwen 3 VL 8B scores 64.6 on olmOCR-bench, below dedicated OCR models |
| Known failure | Sync API takes one page | Can loop and repeat the same text, worse with greedy decoding and on low-quality images. Reported on 10-37% of OCR outputs with transformers 5.17 versus 2-3% baseline. |
Sources
- Textract pricing
- Textract hard limits
- Reducto RD-TableBench
- Qwen3-VL GitHub
- Issue #1611 infinite repetition
- olmo-eval issue #441
What each one is
Amazon Textract
AWS OCR with form key-value pairs, tables and per-word confidence scores.
Amazon Textract detailsQwen-VL
Alibaba's open-weight vision language models (Qwen2.5-VL, Qwen3-VL). Strong general OCR and grounding.
Qwen-VL 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": ["qwen/qwen-vl"],"file_url": "https://example.com/document.pdf"}Questions
More comparisons
Amazon Textract alternativesQwen-VL alternativesBest OCR models in 2026