Mistral Large 4 quickstart

Mistral Large 4 (nicknamed "Le Chonk") has been available as a public preview via Mistral's API since October 6, 2026, under the model identifier mistral-large-4. This page shows the minimal steps to call it: set an environment variable, then send a chat completions request.

1. Prerequisites

• A Mistral API key (the standard Mistral API key). Store it in the MISTRAL_API_KEY environment variable: export MISTRAL_API_KEY="your-key-here"
• Access to the public preview of Mistral Large 4, which began October 6, 2026.
• The base URL is the standard Mistral API endpoint: https://api.mistral.ai/v1/chat/completions.

2. curl example

curl https://api.mistral.ai/v1/chat/completions \
  -H "Authorization: Bearer $MISTRAL_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistral-large-4",
    "messages": [
      {"role": "user", "content": "Explain what a 1M-token context window enables in two sentences."}
    ]
  }'

3. Python example

# pip install requests
import os
import requests

resp = requests.post(
    "https://api.mistral.ai/v1/chat/completions",
    headers={
        "Authorization": f"Bearer {os.environ['MISTRAL_API_KEY']}",
        "Content-Type": "application/json",
    },
    json={
        "model": "mistral-large-4",
        "messages": [
            {"role": "user", "content": "Explain what a 1M-token context window enables in two sentences."}
        ],
    },
    timeout=60,
)
resp.raise_for_status()
print(resp.json()["choices"][0]["message"]["content"])

4. Streaming

The chat completions endpoint also supports streaming responses. Add "stream": true to the request body and iterate over the response stream to receive tokens as they are generated instead of waiting for the full completion.

5. Try without code

If you just want to test the model without writing code, Mistral's docs page (the models page, docs version 26.10) links to a playground where Mistral Large 4 can be tried directly in the browser. See the official Mistral docs for the playground link.

6. Pricing

Public preview pricing is $1.36 per 1M input tokens and $4.18 per 1M output tokens. Full details and comparisons are on the pricing page.

7. Weights and self-hosting

The weights for Mistral Large 4 are not yet available, so self-hosting is not currently possible. See the download page for the current status.