Implementation:Langchain ai Langchain MistralAIEmbeddings

Knowledge Sources	Langchain_ai_Langchain
Domains	Embeddings, MistralAI
Last Updated	2026-02-11 00:00 GMT

Overview

MistralAIEmbeddings is a LangChain embedding model integration that generates text embeddings using the Mistral AI API.

Description

The MistralAIEmbeddings class, defined in the langchain-mistralai partner package, extends both BaseModel (Pydantic) and Embeddings (LangChain core). It generates vector embeddings by sending text to the Mistral AI /embeddings endpoint via httpx clients. The class implements intelligent batching using a tokenizer (the Mixtral-8x7B tokenizer from HuggingFace, or a dummy fallback) to ensure API requests stay within a 16,000-token limit per batch. It supports both synchronous and asynchronous embedding operations, with configurable retry logic using tenacity for handling timeouts and HTTP errors.

Usage

Import this class when you need to generate text embeddings using Mistral AI models for tasks such as semantic search, document similarity, or retrieval-augmented generation (RAG).

Code Reference

Source Location

• Repository: Langchain_ai_Langchain
• File: libs/partners/mistralai/langchain_mistralai/embeddings.py
• Lines: 1-329

Signature

class MistralAIEmbeddings(BaseModel, Embeddings):
    client: httpx.Client = Field(default=None)
    async_client: httpx.AsyncClient = Field(default=None)
    mistral_api_key: SecretStr = Field(alias="api_key", ...)
    endpoint: str = "https://api.mistral.ai/v1/"
    max_retries: int | None = 5
    timeout: int = 120
    wait_time: int | None = 30
    max_concurrent_requests: int = 64
    tokenizer: Tokenizer = Field(default=None)
    model: str = "mistral-embed"

Import

from langchain_mistralai import MistralAIEmbeddings

I/O Contract

Inputs

Name	Type	Required	Description
model	str	No	Name of the Mistral AI embedding model to use. Defaults to "mistral-embed".
mistral_api_key	SecretStr	No	API key for authentication. Read from MISTRAL_API_KEY env var if not provided. Alias: api_key.
endpoint	str	No	Base URL for the Mistral API. Defaults to "https://api.mistral.ai/v1/".
max_retries	int or None	No	Maximum number of retries on failure. Defaults to 5. Set to None to disable.
timeout	int	No	Request timeout in seconds. Defaults to 120.
wait_time	int or None	No	Seconds to wait before retrying on 429 errors. Defaults to 30.
max_concurrent_requests	int	No	Maximum concurrent API requests. Defaults to 64.
tokenizer	Tokenizer	No	HuggingFace tokenizer for batch size calculation. Auto-loaded from mistralai/Mixtral-8x7B-v0.1 or falls back to a dummy tokenizer.

Outputs

Name	Type	Description
embed_documents	list[list[float]]	Returns a list of embedding vectors, one per input document.
embed_query	list[float]	Returns a single embedding vector for a query text.

Usage Examples

Basic Usage

from langchain_mistralai import MistralAIEmbeddings

embed = MistralAIEmbeddings(
    model="mistral-embed",
    # api_key="...",
)

# Embed a single query
vector = embed.embed_query("The meaning of life is 42")
print(vector[:3])

# Embed multiple documents
vectors = embed.embed_documents(["Document 1...", "Document 2..."])
print(len(vectors))

Async Usage

from langchain_mistralai import MistralAIEmbeddings

embed = MistralAIEmbeddings(model="mistral-embed")

vector = await embed.aembed_query("The meaning of life is 42")
print(vector[:3])