Skip to content

Your first Mistral session

At the end of this page you have installed pi-mistral-native, configured an API key, selected the EU region, and sent your first chat message.

  1. Install the package:

    pi install npm:pi-mistral-native

    The provider registers — the model catalog refreshes from Mistral’s live model list.

  2. Log in:

    /login mistral-native

    Paste your API key at the prompt (no echo), then pick the region — choose EU to keep inference inside the EU.

  3. Check the connection:

    /mistral test

    The probe confirms the regional endpoint answers with a model list.

  4. Start a chat — pick a mistral-native/… model and say hello. Reasoning models stream their thinking next to the answer.

Where to go next: