Implementation:Mistralai Client python Chat Stream

Knowledge Sources	Mistral Client Python Mistral AI Docs
Domains	NLP, Streaming, LLM_Inference
Last Updated	2026-02-15 14:00 GMT

Overview

Concrete tool for sending streaming chat completion requests and receiving token-by-token responses via SSE provided by the mistralai SDK.

Description

The Chat.stream() and Chat.stream_async() methods open a streaming HTTP connection to the Mistral API and return an EventStream (sync) or EventStreamAsync (async) iterator. Each iteration yields a CompletionEvent containing a CompletionChunk with incremental response data. The stream must be consumed within a context manager (with / async with) to ensure proper resource cleanup.

Usage

Call client.chat.stream() for synchronous streaming or client.chat.stream_async() with async for for async streaming. Always use within a context manager or ensure the stream is properly closed after consumption.

Code Reference

Source Location

Repository: client-python
File: src/mistralai/client/chat.py
Lines: L462-649 (sync stream), L651-838 (async stream)

Signature

class Chat:
    def stream(
        self,
        *,
        model: str,
        messages: List[ChatCompletionStreamRequestMessage],
        temperature: Optional[float] = None,
        top_p: Optional[float] = None,
        max_tokens: Optional[int] = None,
        stream: Optional[bool] = True,
        response_format: Optional[ResponseFormat] = None,
        tools: Optional[List[Tool]] = None,
        safe_prompt: Optional[bool] = None,
        random_seed: Optional[int] = None,
    ) -> EventStream[CompletionEvent]:
        ...

    async def stream_async(
        self,
        *,
        model: str,
        messages: List[ChatCompletionStreamRequestMessage],
        # Same parameters as stream()
    ) -> EventStreamAsync[CompletionEvent]:
        ...

Import

from mistralai import Mistral
# Access via: client.chat.stream(...)

I/O Contract

Inputs

Name	Type	Required	Description
model	str	Yes	Model ID (e.g., "mistral-large-latest")
messages	List[ChatCompletionStreamRequestMessage]	Yes	Conversation messages
temperature	Optional[float]	No	Sampling temperature
max_tokens	Optional[int]	No	Maximum output tokens
response_format	Optional[ResponseFormat]	No	Output format

Outputs

Name	Type	Description
stream	EventStream[CompletionEvent]	Iterable SSE stream of completion chunks
chunk.data	CompletionChunk	Individual chunk with choices[0].delta.content
chunk.data.choices[0].finish_reason	Optional[str]	Signals stream completion

Usage Examples

Synchronous Streaming

import os
from mistralai import Mistral
from mistralai.models import UserMessage

client = Mistral(api_key=os.environ["MISTRAL_API_KEY"])

with client.chat.stream(
    model="mistral-large-latest",
    messages=[UserMessage(content="Write a short poem about coding.")],
) as stream:
    for chunk in stream:
        content = chunk.data.choices[0].delta.content
        if content:
            print(content, end="", flush=True)
print()  # Final newline

Async Streaming

import asyncio
import os
from mistralai import Mistral
from mistralai.models import UserMessage

async def main():
    client = Mistral(api_key=os.environ["MISTRAL_API_KEY"])

    async with client.chat.stream_async(
        model="mistral-large-latest",
        messages=[UserMessage(content="Explain machine learning.")],
    ) as stream:
        async for chunk in stream:
            content = chunk.data.choices[0].delta.content
            if content:
                print(content, end="", flush=True)
    print()

asyncio.run(main())

Related Pages

Implements Principle

Principle:Mistralai_Client_python_Streaming_Chat_Completion

Requires Environment

Environment:Mistralai_Client_python_Python_SDK_Environment

Uses Heuristic

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment