An open-source, minimal Python wrapper around the Mistral API focused on OCR use cases where visual structure matters—such as tables, charts, and diagrams. The wrapper produces markdown output with image clippings embedded directly, making it suitable for downstream documentation and knowledge pipelines.

Links

Project Repository | Medium Article

Key Features

• Architected a minimal Python wrapper that completely abstracts away verbose Mistral SDK setup and boilerplate.
• Optimized for complex document structures, focusing heavily on scenarios where visual context (tables/charts) must be preserved alongside scraped text.
• Engineered automatic Base64 image embedding, directly injecting extracted image clippings into the unified markdown output.
• Designed a deliberately small API surface, exposing only two simple asynchronous entry points for rapid developer integration.

# OCR an image
await mistral_client.ocr_image("path/to/image.png")

# OCR a PDF
await mistral_client.ocr_pdf("path/to/document.pdf")