AGPL-3.0-or-later · forever.

the provider catalog

Mistral AI

Frontier models from Paris: a European home for workflows that must stay under European law, with open-weight releases alongside the API. One provider/model line moves a workflow here or away again: the same file runs on every provider the catalog names, and swapping is a one-word diff. Machines read the catalog; the binary answers nika catalog.

  • 2pinned modelsaliases pin exact upstream ids
  • cloudfamilybring your own key; it rides an env var and stays yours
  • MISTRAL_API_KEYthe key rides herean env var, never a config file
  • openai-chatwire dialectthe adapter speaks it

the wire

cloud · bring your own key; it rides an env var and stays yours

A workflow names this provider in the model line (model: mistral/mistral-large-latest), and the engine speaks openai-chat on the wire. The key rides MISTRAL_API_KEY in the environment, never a config file, never the workflow.

default
mistral-large-latest
cheap
mistral-small-latest
dialect
openai-chat
tags
code · european · frontier · function-calling · open-source · streaming · structured-output

the models

2 pins

Aliases are what the file may write; the engine pins each one to an exact upstream id, so a run stays reproducible after a vendor rotates names.

  • largemistral-large-latestctx 128k · vision · json:schema$0.5/M in · $1.5/M out
  • smallmistral-small-latestctx 32k · vision · json:schema$0.15/M in · $0.6/M out

prices are the ENGINE's, not ours: the audit's pricing snapshot (models.dev · pinned 2026-07-07): the same table nika check prices your ceiling with.

in the released catalog

2 rooms

The wire models this provider serves in the released binary's catalog · each one is a room with its seats, exact prices and measured energy.

the whole register: /catalog/models · what the released binary knows, vendored at the engine pin.

in a real file

a complete workflow · shown whole
draftinfer
translateinfer
mistral.nika.yaml
nika: v1workflow:  id: bilingual-release-note  description: "drafted in French, shipped in English — large drafts, small translates"tasks:  draft:    infer:      model: mistral/mistral-large-latest      max_tokens: 120      prompt: |        Rédige en une phrase la note de version : les traces        sont désormais signées et rejouables.  translate:    with:      draft: ${{ tasks.draft.output }}    infer:      model: mistral/mistral-small-latest      max_tokens: 120      prompt: |        Translate to plain English: ${{ with.draft }}outputs:  note: ${{ tasks.translate.output }}

check-green against the released engine · 2 tasks · 2 waves · worst-case ceiling $0.0003 · the drift gate re-proves this copy on every test run · open it in the playground →

get set up

keys at console.mistral.ai
  1. 1Create a key in the console (API Keys).
  2. 2The key rides the environment, never the file export MISTRAL_API_KEY=…

Pick it for when European jurisdiction is a requirement, not a preference. Mistral AI · Paris, France. Then prove the seat: nika doctor names every provider this binary can reach.

swap it

10 more cloud seats

Moving off Mistral AI is editing one line. The rest of the file (tasks, permits, outputs, the trace) does not change. The register orders the catalog local first; beyond it the engine embeds 21 more OpenAI-compatible endpoints.

Try it in the playground, or install and go local first. Read the spec →