Mistral OCR
Mistral OCR is a document extraction model from Mistral designed to recognize and extract text from images and documents.
Document and image text extraction via OCR
Mistral OCR is a document extraction model developed by Mistral, a French AI company. It is designed to process images and documents and extract structured text from them using optical character recognition techniques. The model is available via Mistral's API under the identifier mistral-ocr-latest, reflecting that it is updated as the canonical latest version of the OCR offering.
Mistral OCR is best suited for workflows that require converting scanned documents, PDFs, or image-based files into machine-readable text. It fits into pipelines where downstream processing — such as search indexing, data extraction, or document analysis — depends on accurate text recognition. Because it is a first-party model served directly by Mistral, it integrates naturally with other Mistral API capabilities.
What Mistral OCR supports
Text Extraction
Extracts machine-readable text from images and document files using optical character recognition. Outputs structured text suitable for downstream processing.
Document Processing
Handles document-type inputs such as scanned PDFs and image-based files, converting their content into extractable text.
Image Reading
Reads and interprets text embedded within images, supporting use cases like digitizing printed or handwritten materials.
API Integration
Accessible via Mistral's first-party API using the model ID mistral-ocr-latest, enabling integration into existing application pipelines.
Ready to build with Mistral OCR?
Get Started FreeCommon questions about Mistral OCR
What is the context window for Mistral OCR?
The context window for Mistral OCR is not specified in the available metadata. Because it is a document extraction model rather than a generative language model, context window limits may not apply in the traditional sense.
How is Mistral OCR priced?
Pricing information for Mistral OCR is not included in the available metadata. You should check Mistral's official pricing page at mistral.ai for current rates.
What model ID do I use to call Mistral OCR via the API?
The model ID is mistral-ocr-latest. This identifier always points to the most current version of Mistral's OCR model.
What types of inputs does Mistral OCR accept?
Mistral OCR is designed to process images and document files such as scanned PDFs. It extracts text content from these inputs using optical character recognition.
Does Mistral OCR have a knowledge cutoff date?
Mistral OCR is a document extraction model, not a generative language model, so a knowledge cutoff date is not applicable. It processes the content of documents provided to it at inference time.
Explore similar models
MistralMinistral 3 3B
Ministral 3 3B is a 3-billion-parameter open-source text generation model from Mistral with a 256,000-token context window.
MistralMinistral 3 8B
Ministral 3 8B is an open-source text generation model from Mistral with a 256,000-token context window.
MistralMinistral 3 14B
Ministral 3 14B is an open-source text generation model from Mistral with a 256,000-token context window.
GoogleGoogle Document AI
Google Document AI is a cloud-based document extraction service from Google that uses OCR to parse and structure document content.
LlamaIndexLlamaParse
LlamaParse is a document extraction service by LlamaIndex designed to parse complex documents for use in LLM pipelines.
ByteDanceSeedance 2.5 Turbo
Seedance 2.5 Turbo is a video generation model from ByteDance that supports text-to-video, image-to-video, and video editing modes.
Start building with Mistral OCR
No API keys required. Create AI-powered workflows with Mistral OCR in minutes — free.