Mistral (LLM)

Use Mistral as the LLM (dialogue + reasoning) provider.

Setup

• Set MISTRAL_API_KEY

# .env
MISTRAL_API_KEY=...

Example

from siphon.plugins import mistralai

llm = mistralai.LLM(
    model="mistral-large-latest",
)

Common options

• model (default: mistral-medium-latest)
• api_key (default: None)
• temperature (default: 0.3)
• max_completion_tokens (default: 150)
• timeout (default: 15)