Mistral AI
Chat completions, embeddings, and OCR via the Mistral AI API.
About
Mistral AI ships in the w6w first-party pack. It declares 4 actions, 2 health checks, and the host runs its code in a sandbox that never sees the credential.
- App id
io.w6w.mistral- Version
- 0.3.1
- Author
- w6w
- Licence
- MIT
- Categories
- AI & Machine Learning
Overview
Mistral runs chat completions and embeddings from a workflow — sending a chat-completion request to any available model, and generating embeddings for text so a workflow can do semantic search or similarity matching downstream.
Text extraction covers Mistral’s OCR capability directly: documents can be turned into text as an action in their own right, rather than a workflow having to route through a separate OCR service. Models can be listed to keep a workflow’s model choice explicit and current rather than hardcoded against whatever was available when it was built.
A fit for adding an LLM step to a workflow — summarizing or classifying text, extracting text from a scanned document, or generating embeddings to feed a search or recommendation step elsewhere in the process.
Build with Mistral AI
Three routes to the same 4 actions. The Workflow tab is generated from Mistral AI's own manifest and carries its real ids, so it is copy-pasteable; the Code and CLI examples are the same call for any action on any app, so every app-specific value in them is a blank you fill in.
A workflow step names the app and the action, and the editor fills in the
connection when you pick one. This is the Step shape from the
workflow spec, carrying Mistral AI's real ids.
{
"manifestVersion": "2",
"name": "mistral-example",
"steps": [
{
"id": "chat-completion",
"uses": {
"app": "io.w6w.mistral",
"action": "chat-completion",
"connection": "conn_YOUR_CONNECTION_ID"
},
"with": {
"model": "<model>",
"messages": "<messages>"
}
}
]
}Here are some of the things you can do
-
Chat Completion
performchat-completion -
Create Embeddings
performembeddings -
Extract Text (OCR)
performextract-text -
List Models
readlist-models
Every app-specific value here is a blank you have to fill in. An
app action is reached through the connection that authenticates it, so the
address is a connection id, not the app id — and connections belong to your account,
so a public page cannot know yours. Create one for Mistral AI, then fill in
the three blanks: conn_YOUR_CONNECTION_ID, the action key, and the
parameters that action declares. The call itself is real — the shape is transcribed
from the studio's own snippet builder, which prints the same kind of blanks — but
nothing in it is specific to Mistral AI. The Workflow tab is where this app's
real ids are.
npm install @w6w/sdkyarn add @w6w/sdkpnpm add @w6w/sdkdeno add npm:@w6w/sdkimport { W6wClient, isActionRun } from "@w6w/sdk";
// Reads W6W_BASE_URL and W6W_TOKEN from the environment when omitted.
const client = new W6wClient();
const envelope = await client.run({
urn: "conn_YOUR_CONNECTION_ID",
action: "chat-completion",
payload: {
model: "<value>",
messages: "<value>",
// temperature: "<value>",
// topP: "<value>",
// maxTokens: "<value>",
// stop: "<value>",
// randomSeed: "<value>",
// responseFormat: "<value>",
// safePrompt: "<value>",
// tools: "<value>",
// toolChoice: "<value>",
// parallelToolCalls: "<value>",
},
});
if (isActionRun(envelope)) console.log(envelope.value);npm install -g @w6w/cliw6w run conn_YOUR_CONNECTION_ID --action chat-completion --payload '{"model":"<value>","messages":"<value>"}'Give an AI agent Mistral AI — without giving it Mistral AI's credentials. One MCP endpoint exposes every app, function and workflow the caller is entitled to, as tools it can discover and run. Access is granted per team while we onboard.
One tool call{
"name": "w6w_invoke",
"arguments": {
"ref": "app:io.w6w.mistral#chat-completion",
"input": {
"model": "<model>",
"messages": "<messages>"
}
}
}
Every tool names its target with a single ref. The
app: form above doesn't name a connection at all — the
host resolves which of the caller's Mistral AI connections to sign
with, and refuses rather than guesses when the answer is ambiguous.
What the agent gets
Credentials it can't read
The token is attached host-side, at the moment of the call. It is never a tool argument, never in the model's context, and never in a transcript — so a prompt injection has nothing to exfiltrate.
A tool surface scoped to the caller
Tools are derived per end user from what that person has actually connected and is entitled to — not one shared bot identity carrying the union of everyone's access.
A durable workflow in one call
Multi-step work runs on the workflow engine and returns a run handle the agent can poll — retries, branching and state survive the conversation that started them.
Health-aware discovery
Mistral AI's declared health checks are on the surface too, so an agent can tell "the vendor is down" from "your credential expired" before it burns a retry on either.
The MCP surface is part of the hosted platform. Mistral AI itself is MIT, and the runtime that executes it is source-available (FSL).
Health checks
Mistral AI declares its own checks, so its health is a property of the app rather than something the host guesses at.
Mistral platform status
Open incidents from Mistral's status feed. The host folds updates per incident; state comes from each incident's `Status:` field, not from the newest headline.
API rate-limit headroom
Per-minute request and token allowances remaining, read off the response headers.