OCR Mistral

[Hekas1]

good day, can you please tell me if it is possible to use Mistral-ocr to extract PDF text, using this model I get the following error:
[ERROR]Internal server error while chunking: Cant find model for 7edfadd64e0f11f0be09bebabe11560d/image2text/mistral-ocr-latest@Mistral 02:40:47 [ERROR][Exception]: Can't find model for 7edfadd64e0f11f0be09bebabe11560d/image2text/mistral-ocr-latest@Mistral

However, in another interface this extraction method works properly

[dosubot[bot]]

RAGFlow does not natively support Mistral-ocr for PDF text extraction, even though mistral-ocr-latest appears in the configuration file. The error happens because there is no implementation or registration for Mistral-ocr as an image2text/OCR model in the codebase, so RAGFlow can't instantiate or use it at runtime. To use Mistral-ocr, you would need to implement and register it in the codebase similarly to other supported image2text models (like Gemini, Ollama, etc.)—just adding it to the config is not enough for it to work in RAGFlow. This is why it works in another interface but fails here: the other interface likely has native support or a plugin for Mistral-ocr, while RAGFlow currently does not see details.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}