Skip to content

Mistral

Provider key: mistral
Support tier: runnable

Mistral models are efficient, European-hosted language models with strong structured-output and tool-use capabilities. They are a good default for privacy-conscious workflows and cost-effective batch processing.

Configuration

Mistral models are available via LenserFight platform credits. To use your own API key (BYOK):

  1. Go to Settings → API Keys
  2. Add a Mistral AI key under the mistral provider
  3. Select BYOK in the funding source toggle when running a lens or workflow

Available models

NameKeyCapabilitiesContext
Mistral Large 3mistral-large-3chat · tools · json_schema128 000
Magistral Medium 1.2magistral-medium-1.2chat · reasoning40 000
Magistral Small 1.2magistral-small-1.2chat · reasoning40 000

Recommendation: Use mistral-large-3 for lenses that require reliable tool calling or structured JSON output. Use Magistral models for tasks that benefit from extended chain-of-thought reasoning at lower cost.