​
Requirements

mistralai
maxim-py

​
Env variables

MAXIM_API_KEY=
MAXIM_LOG_REPO_ID=
MISTRAL_API_KEY=

​
Initialize logger

from maxim import Maxim

logger = Maxim().logger()

​
Initialize MaximMistralClient

from mistralai import Mistral
from maxim.logger.mistral import MaximMistralClient
import os

with MaximMistralClient(Mistral(
    api_key=os.getenv("MISTRAL_API_KEY", ""),
), logger) as mistral:
    # Your Mistral calls go here
    pass

​
Make LLM calls using MaximMistralClient

from mistralai import Mistral
from maxim.logger.mistral import MaximMistralClient
import os

with MaximMistralClient(Mistral(
    api_key=os.getenv("MISTRAL_API_KEY", ""),
), logger) as mistral:

    res = mistral.chat.complete(
        model="mistral-small-latest", 
        messages=[
            {
                "content": "Who is the best French painter? Answer in one short sentence.",
                "role": "user",
            },
        ]
    )

    # Handle response
    print(res)
mistral_traces.gif

​
Async LLM calls

async with MaximMistralClient(Mistral(
    api_key=os.getenv('MISTRAL_API_KEY', ''),
), logger) as mistral:
    
    response = await mistral.chat.complete_async(
        model='mistral-small-latest',
        messages=[
            {
                'role': 'user',
                'content': 'Explain the difference between async and sync programming in Python in one sentence.'
            }
        ]
    )    
    print(response)

​
Supported Mistral Models

The MaximMistralClient supports all Mistral models available through the Mistral API, including:
  • mistral-small-latest
  • mistral-medium-latest
  • open-mistral-7b
  • And other available Mistral models

​
Resources

You can quickly try the Mistral One Line Integration here -

Mistral Integration Notebook (Colab)