Mistral: Pixtral 12B

Model Overview

Description

The first multi-modal, text+image-to-text model from Mistral AI. Its weights were launched via torrent, making it openly available for research and commercial use.

Pixtral 12B represents Mistral AI's entry into the vision-language model space, combining strong text understanding with image comprehension capabilities.

Specifications

Pricing

Via Hyperbolic Provider

Capabilities

• Multi-modal Input: Accepts both text and image inputs
• Text Generation: Produces text-only outputs
• Standard Parameters: Supports temperature, top-p, frequency penalty, and presence penalty
• Tool Calling: Supported
• Reasoning: Not available (standard generation model)

Supported Parameters

Use Cases

• Image Analysis: Describe, analyze, and extract information from images
• Document Understanding: Process documents with text and visual elements
• Visual Q&A: Answer questions about image content
• Content Moderation: Analyze images for content classification
• Accessibility: Generate alt-text descriptions for images
• General Chat: Standard conversational AI tasks

Provider Information

Primary Provider: Hyperbolic

Usage Examples

Basic Text Completion

curl -X POST https://api.langmart.ai/v1/chat/completions \
  -H "Authorization: Bearer $LANGMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistralai/pixtral-12b",
    "messages": [
      {
        "role": "user",
        "content": "Explain the concept of neural networks in simple terms."
      }
    ]
  }'

Image + Text Input (Vision)

curl -X POST https://api.langmart.ai/v1/chat/completions \
  -H "Authorization: Bearer $LANGMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistralai/pixtral-12b",
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "text",
            "text": "What do you see in this image?"
          },
          {
            "type": "image_url",
            "image_url": {
              "url": "https://example.com/image.jpg"
            }
          }
        ]
      }
    ]
  }'

Python SDK Example

from openai import OpenAI

client = OpenAI(
    base_url="https://api.langmart.ai/v1",
    api_key="your-langmart-api-key"
)

# Text-only request
response = client.chat.completions.create(
    model="mistralai/pixtral-12b",
    messages=[
        {"role": "user", "content": "What is machine learning?"}
    ],
    temperature=0.7,
    max_tokens=500
)

print(response.choices[0].message.content)

Vision Request with Python

from openai import OpenAI
import base64

client = OpenAI(
    base_url="https://api.langmart.ai/v1",
    api_key="your-langmart-api-key"
)

# With image URL
response = client.chat.completions.create(
    model="mistralai/pixtral-12b",
    messages=[
        {
            "role": "user",
            "content": [
                {"type": "text", "text": "Describe this image in detail."},
                {
                    "type": "image_url",
                    "image_url": {"url": "https://example.com/photo.jpg"}
                }
            ]
        }
    ]
)

print(response.choices[0].message.content)

Comparison with Similar Models

Notes

• Weights are openly available via torrent
• First multi-modal model from Mistral AI
• Optimized for efficient inference with moderate parameter count
• Well-suited for applications requiring both text and image understanding

References

• LangMart Model Page
• Mistral AI
• Hyperbolic API

Last updated: December 2024