AlibabaQwen OCR
Qwen3.5-OCR
Qwen3.5-OCR adds multi-turn conversations and PDF parsing to Alibaba's OCR model line. It is based on Qwen3.5.
Technical details
The Responses interface accepts PDFs. DashScope exposes additional built-in OCR controls.Qwen OCR guide
Practical use
Practical guidance. Use the document structure to connect every extracted value to its source. Check rotated pages, small type and merged table cells with your real document set.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | API |
| Model identifiers | qwen3.5-ocr |
| Document support | Text positions, document parsing and key-field extraction.Qwen OCR guide |
Related models
- Qwen-VL-OCR
Alibaba ยท Qwen OCR
Qwen3-VL-based OCR model for text localization, table and formula recognition, and image rotation correction.
Launch date not confirmedAPI