Mistral AI Embeddings

Generate high-quality embeddings using Mistral AI's advanced language models. Features robust rate limiting, automatic retries, and configurable API settings for enterprise deployments.

Mistral AI Embeddings component interface and configuration

API Key Notice: A valid Mistral AI API key is required to use this component. Ensure your API key has sufficient quota and appropriate rate limits for your embedding needs.

Component Inputs

Model: The Mistral AI embedding model to use
Example: "mistral-embed", "mistral-embed-light"
Mistral API Key: Your Mistral AI API authentication key
Example: "m-abc123xyz456..."
Max Concurrent Requests: Maximum number of parallel requests
Example: 5 (Default)
Max Retries: Number of retry attempts for failed requests
Example: 3 (Default)
Request Timeout: Timeout in milliseconds for API requests
Example: 30000 (Default, 30 seconds)
API Endpoint: Custom API endpoint URL
Example: "https://api.mistral.ai/v1" (Default)

Component Outputs

Embeddings: Vector representation of the input text
Example: [0.018, -0.032, 0.067, ...]
Dimensions: The dimension size of the embedding vector
Example: 1024 for mistral-embed, 512 for mistral-embed-light
Usage: Token usage information for billing purposes
Example: prompt_tokens: 12, total_tokens: 12

Model Comparison

mistral-embed

Mistral AI's high-performance embedding model with superior semantic understanding

Dimensions: 1024
Performance: High-quality semantic representations
Language Support: Multilingual
Ideal for: Production systems requiring top-tier semantic search and retrieval

mistral-embed-light

Lightweight and efficient embedding model with faster processing

Dimensions: 512
Performance: Faster processing with good quality
Language Support: Multilingual with focus on major languages
Ideal for: Cost-effective embeddings at scale or latency-sensitive applications

Implementation Example

// Basic configuration
const embedder = new MistralAIEmbeddor({
  model: "mistral-embed",
  mistralApiKey: process.env.MISTRAL_API_KEY,
});

// Advanced configuration
const advancedEmbedder = new MistralAIEmbeddor({
  model: "mistral-embed-light",
  mistralApiKey: process.env.MISTRAL_API_KEY,
  maxConcurrentRequests: 10,
  maxRetries: 5,
  requestTimeout: 60000,
  apiEndpoint: "https://custom-endpoint.com/v1"
});

// Generate embeddings
const result = await embedder.embed({
  input: "Your text to embed"
});

// Batch processing
const batchResult = await embedder.embedBatch({
  inputs: [
    "First text to embed",
    "Second text to embed"
  ]
});

console.log(result.embeddings);

Use Cases

Enterprise Search: Build high-quality semantic search for document retrieval
Retrieval-Augmented Generation: Create advanced RAG systems with precise context retrieval
Content Recommendation: Develop intelligent content recommendation engines
Multilingual Applications: Support global content with strong multilingual capabilities
Hybrid Search Solutions: Combine with keyword search for comprehensive retrieval systems

Useful Resources

Best Practices

Use environment variables for API keys in production environments
Implement caching for frequently used embeddings to reduce API costs
Monitor rate limits and adjust concurrent requests accordingly
Choose mistral-embed-light for high-volume or latency-sensitive applications
Set appropriate timeout values based on your application's requirements

Documentation