Blog

Mistral OCR Is Now Available in Puter.js

Nariman Jelveh

We're excited to announce that Puter.js now supports Mistral OCR, giving developers free, unlimited access to one of the most advanced optical character recognition systems available today.

What is Mistral OCR?

Mistral OCR is a state-of-the-art text extraction service that goes far beyond traditional OCR capabilities. Built by Mistral AI, it combines deep learning with document understanding to deliver exceptional accuracy across multiple languages, complex layouts, and diverse content types.

What makes Mistral OCR special:

  • Superior multilingual support: Automatically handles documents in 100+ languages without configuration
  • Document structure understanding: Preserves formatting, tables, and layout information
  • Advanced annotations: Optional bounding box data for precise text positioning
  • Complex layout handling: Excels at multi-column text, tables, and mixed content
  • Handwriting recognition: Industry-leading accuracy for handwritten text

Through Puter.js, you get all of this completely free—no Mistral API key, no usage limits, no backend setup required.

Quick Examples

Basic text extraction:

puter.ai.img2txt({
    source: 'https://example.com/document.jpg',
    provider: 'mistral'
});

Multilingual document processing:

// Automatically detects and processes any language
puter.ai.img2txt({
    source: multilingualDocument,
    provider: 'mistral',
    model: 'mistral-ocr-latest'
});

Advanced options with annotations:

puter.ai.img2txt({
    source: document,
    provider: 'mistral',
    bboxAnnotationFormat: 'yolo',
    includeImageBase64: true
});

Why Mistral OCR?

While Puter.js already supported AWS Textract (which remains the default provider), Mistral OCR brings several key advantages:

Feature AWS Textract Mistral OCR
Basic text extraction ✓ ✓
Multilingual support Good Exceptional
Document annotations Limited Rich & detailed
Handwriting accuracy Good Industry-leading
Complex layouts Good Excellent
Format preservation Basic Advanced

Mistral OCR is particularly powerful for:

  • Processing documents in multiple languages
  • Extracting structured data from forms and tables
  • Digitizing historical documents with varied handwriting
  • Building applications that need precise text positioning
  • Handling mixed-content documents (text + images + diagrams)

Get Started Now

Just add one script tag to your HTML:

<script src="https://js.puter.com/v2/"></script>

Then use Mistral OCR with a single function call:

const text = await puter.ai.img2txt({
    source: yourImage,
    provider: 'mistral'
});

That's it. No sign-ups, no API keys, no server infrastructure. Start extracting text from images immediately with one of the most advanced OCR systems available.

Complete Demo

Here's a working example that shows Mistral OCR processing a multilingual document:

<html>
<body>
    <script src="https://js.puter.com/v2/"></script>
    
    <div style="display: flex; gap: 20px; align-items: flex-start;">
        <div style="flex: 1;">
            <img src="https://assets.puter.site/letter.png" style="max-width: 100%; height: auto;">
        </div>
        <div style="flex: 1;">
            <div id="result" style="white-space: pre-wrap; border: 1px solid #ccc; padding: 10px; min-height: 200px;">Extracting text...</div>
        </div>
    </div>

    <script>
        const resultDiv = document.getElementById('result');
        
        puter.ai.img2txt({
            source: 'https://assets.puter.site/letter.png',
            provider: 'mistral'
        }).then(response => {
            resultDiv.textContent = response || 'No text found';
        })
    </script>
</body>
</html>



That's it! No sign-ups, no API keys, no server infrastructure required. Start extracting text from images immediately with one of the most advanced OCR systems available.


Learn more:

Free, Serverless AI and Cloud

Start creating powerful web applications with Puter.js in seconds!

Get Started Now

Read the Docs • Try the Playground