Mistral

Usage

from tokenid import JournaledMistral

client = JournaledMistral(
    api_key="…",
    session_id="my-session",
)

resp = client.chat.complete(
    model="mistral-large-latest",
    messages=[{"role": "user", "content": "Hello"}],
)

JournaledMistral accepts every argument that mistralai.Mistral accepts and forwards every method.

Streaming

for chunk in client.chat.stream(
    model="mistral-large-latest",
    messages=[{"role": "user", "content": "Hello"}],
):
    ...

Tokens are aggregated across the final stream chunk and journaled on completion.

Supported models

Pricing is built in for Mistral Large, Small, Nemo, and Codestral. Cost is computed from response.usage.prompt_tokens and response.usage.completion_tokens.

Historical usage

Mistral does not expose a historical usage API — only response-level usage at call time. You must use the SDK wrapper to capture usage; there is no backfill path for historical Mistral data.