Vision

Vision capabilities enable models to analyze images and provide insights based on visual content in addition to text. This multimodal approach opens up new possibilities for applications that require both textual and visual understanding.

We provide a variety of models with vision capabilities, all available via the Chat Completions API.

Models with Vision Capabilities

• Pixtral 12B via pixtral-12b-latest
• Pixtral Large via pixtral-large-latest
• Mistral Medium 3.1 via mistral-medium-2508
• Mistral Small 3.2 via mistral-small-2506

Use Vision Models

There are two ways to send an image to the Chat Completions API, either by passing a URL or by passing a base64 encoded image.

import os
from mistralai import Mistral

api_key = os.environ["MISTRAL_API_KEY"]
model = "mistral-small-2506"

client = Mistral(api_key=api_key)

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "text",
                "text": "What's in this image?"
            },
            {
                "type": "image_url",
                "image_url": "https://docs.mistral.ai/img/eiffel-tower-paris.jpg"
            }
        ]
    }
]

chat_response = client.chat.complete(
    model=model,
    messages=messages
)

import os
from mistralai import Mistral

api_key = os.environ["MISTRAL_API_KEY"]
model = "mistral-small-2506"

client = Mistral(api_key=api_key)

messages = [
    {
        "role": "user",
        "content": [
            {
                "type": "text",
                "text": "What's in this image?"
            },
            {
                "type": "image_url",
                "image_url": "https://docs.mistral.ai/img/eiffel-tower-paris.jpg"
            }
        ]
    }
]

chat_response = client.chat.complete(
    model=model,
    messages=messages
)

Below you can find a few examples of use cases leveraging our models vision, from understanding graphs to extract data, the use cases are diverse.

Charts

Compare images

Transcribe

OCR Old Documents

OCR with Structured output

¡Meow! Click one of the tabs above to learn more.