PaddleOCR API

About the project

FastAPI server for image text extraction using PaddleOCR v5. Processes images directly from URLs without file uploads, supports multi-language OCR (Spanish, English and more), confidence-based filtering, and auto-cleanup of temporary files. Includes full test coverage with pytest and httpx.

Highlights

• Processes images directly from URLs — no file uploads needed
• Multi-language OCR with confidence-based filtering
• Full test coverage with pytest and httpx

Built with