Resources · Make integration

Mistral AI Make integrationAutomate Mistral AI with Make.

The Mistral AI Make integration gives you 4 modules and no code to write. A module is one brick of the scenario: three of them act, one searches. If you have never built a scenario, an automation made of chained modules, this page walks you from the connection to your first run.

Try the Mistral AI integration on MakeSee the 4 modules

Verified Trustpilot reviews · AI, automation & growth agency

What you can automate

What can the Mistral AI Make integration really do?

The Mistral AI Make integration lets a scenario call your Mistral AI account on its own: Make sends the text, Mistral AI answers, and that answer keeps travelling to the next module. You set it up once with your API key, and all four modules of the app reuse that connection, your Mistral AI account linked to Make.

Three needs come up again and again, and each one leans on a different module. Need a written answer right away? Create a Chat Completion takes your instruction plus the data already moving through the scenario and writes it for you. Need to search a pile of text by meaning instead of by exact wording? Create Embeddings turns each piece of text into a series of numbers a database can compare. And before building anything on top of a connection, List Models checks with your account which models, the versions of Mistral AI it can call, it is allowed to use.

What Make will not do with this app: Mistral AI has no trigger module. Nothing happening on the Mistral AI side can wake a scenario up. So a scenario that uses Mistral AI starts on a schedule, the clock sitting on its first module, or on the trigger of another app, and calls Mistral AI in the middle of the chain.

One more caveat, and it comes from the documentation itself: the app page published on Make's side is generated from Mistral AI's own docs and can contain mistakes. Anything that matters gets checked against Mistral AI.

The same app also exists as an n8n node. Choosing between the two tools has more to do with where you want to host and maintain your automations than with this integration: the n8n vs Make comparison goes through the criteria, and the Mistral AI n8n integration page lists what the node covers. If the scenario should be built for you rather than by you, the Make agency page explains how the team works.

Connect

How do you connect your Mistral AI account to Make?

  1. 01

    Create your API key

    In your Mistral AI workspace, open the API keys page and click Create new key. The key shows up on screen: copy it right then, that is the step everyone skips. Make will ask nothing else of your Mistral AI account.

  2. 02

    Add a Mistral AI module

    Over in Make, open a scenario, add any Mistral AI module and click Create a connection. A connection is your account of the app linked to Make once and for all, and every other Mistral AI module finds it on its own afterwards.

  3. 03

    Paste your credentials

    The Connection name field is there to label the connection when you juggle several accounts, and it stays optional. Paste the key you copied, follow what the window asks for, and the module is ready to be filled in.

First scenario

Your first scenario, step by step

GoalWhen the scenario runs, Mistral AI writes the answer and Make hands it to the next module.

  1. 01

    Create the scenario

    From the Scenarios page, create a new scenario and click the + to place your first module. Nothing runs yet: the editor is a blank canvas waiting for its first brick.

  2. 02

    Place Create a Chat Completion

    Search Mistral AI by name, pick Create a Chat Completion, and link your account with Create a connection if that is not done yet. The module lands on the canvas, still empty.

  3. 03

    Write what you expect

    Fill the fields of the module with the request you would hand to a colleague. Keep the first attempt short: one clear question is enough to see the shape of what comes back.

  4. 04

    Add the second module

    Click the + on the right of the module and add Create Embeddings, then pull in the text the completion returned. That link between two modules is what makes a chain rather than two separate calls.

  5. 05

    Test with Run once

    Click Run once. The scenario runs a single time and each module shows the bundles it received, a bundle being one item passing from a module to the next. Read them before going further.

  6. 06

    Schedule it and switch it on

    Set the schedule on the first module. A new scenario is set to run every 15 minutes, so widen that interval if it fires more often than you need, then flip the scenario on.

Modules

What each Mistral AI module does

Mistral AI gives you 4 modules. For each one: what it does for you, when to reach for it, and what to watch out for.

Mistral AI1

Create a Chat Completion

Action

In Make“Creates a chat completion.”

This is the module that makes Mistral AI write. You hand it your instruction plus the data already sitting in the scenario, and it returns the answer as text the next module can pick up.

When to use it
a support team that wants every long message summed up in three lines before a human opens it.
Watch out
each run counts as one operation, the unit Make deducts from your plan, so test with Run once before switching a busy scenario on.
Mistral AI2

Create Embeddings

Action

In Make“Create embeddings for given text.”

Create Embeddings converts a piece of text into the series of numbers a search engine or a database knows how to compare. Text goes in, that representation comes out, and the rest of the scenario stores it or matches it.

When to use it
a knowledge base you want to query by meaning instead of by exact wording.
Watch out
it produces the representation and nothing more, so decide which module will store it before you build this one.
Mistral AI3

Make an API Call

Action

In Make“Performs an arbitrary authorized API call.”

The catch-all module: it sends the request you write to the Mistral AI API, using the connection already in place. Everything the three other modules do not cover goes out through here.

When to use it
a capability described in Mistral AI's documentation that has no module of its own in Make.
Watch out
nothing is prepared for you here, so keep Mistral AI's documentation open next to the editor.
Mistral AI4

List Models

Search

In Make“Retrieves a list of available models to use.”

List Models asks your account which models are available and hands them back to the scenario. Useful as a first module when you want proof the connection answers before building anything on top of it.

When to use it
when you are breaking in a fresh connection, or when the scenario should choose a model instead of carrying one written in stone.
Watch out
a search module returns as many bundles as it finds items, so whatever sits behind it runs once per model.
Need help

Need help automating Mistral AI with Make?

A person reads every message.

FAQ

Questions that come up next

01Is the Mistral AI Make integration free?
On Make's side, yes. Mistral AI is a standard app, so it comes with every plan, Free included, and Make bills nothing extra to use it. The Free plan sets the frame: 2 active scenarios, 15 minutes minimum between two scheduled runs, 5 minutes of run time per execution and 512 MB of data transfer. Paid plans bring the interval down to 1 minute, lift the cap on active scenarios and stretch run time to 40 minutes. What Mistral AI charges on its own side is a separate question, answered on its pricing page.
02What do you need to connect Mistral AI to Make?
An active Mistral AI account and an API key, nothing more. The key is created in your Mistral AI workspace, on the API keys page, with Create new key, and it gets copied while it is displayed on screen. In Make you add a Mistral AI module, click Create a connection, paste the key and give the connection a name if several accounts cross paths. That name stays optional. Once the connection exists, every Mistral AI module you add later reuses it, which makes this a one time job.
03Can Mistral AI fire a scenario in real time?
No, and it is not a question of delay: Mistral AI has no trigger module in Make. Nothing leaves Mistral AI on its own initiative. Two ways to start anyway. Either the schedule, the clock on the first module, which a brand new scenario sets to every 15 minutes. Or the trigger of another app, with the Mistral AI module placed further down the chain. The second one is what most scenarios end up doing: something arrives, Mistral AI reads it, the answer carries on.
04What if the module you need is missing?
That is what Make an API Call is for. It sends the request you define to the Mistral AI API with the same connection, so a capability without a dedicated module stays within reach of the scenario. Prepare it with Mistral AI's documentation in front of you: nothing is pre-filled, the call is yours to write. Worth knowing too, the app page published on Make's side was generated from Mistral AI's documentation and can contain mistakes, so anything important gets verified at the source.
05Make or n8n for Mistral AI?
The honest answer depends on what your team already runs, not on which tool is better at Mistral AI. Make favors speed: build the scenario visually, link the account with a single key, and watch each bundle land with Run once. n8n fits teams that host their own automations and want to stay close to code, and it carries the same Mistral AI node. Wiring up the app takes about the same effort either way; the real question is which editor your team will keep open next month.
Hack'celeration Lab

Get our weekly automation tips.

No spam. Unsubscribe anytime.

Same platform

Other docs you might need on this stack.

Browse all integrations