Skills Artificial Intelligence Mistral AI Chat Completion Quickstart

Mistral AI Chat Completion Quickstart

v20260423
mistral-hello-world
This comprehensive guide provides minimal working examples for integrating with the Mistral AI API. It demonstrates core LLM functionalities including basic chat completion, streaming responses, multi-turn conversations, and generating structured JSON output. Use this guide when starting a new Mistral integration, testing API connections, or learning advanced AI patterns using TypeScript or Python.
Get Skill
148 downloads
Overview

Mistral AI Hello World

Overview

Minimal working examples demonstrating Mistral AI chat completions, streaming, multi-turn conversation, and JSON mode. Uses the official @mistralai/mistralai TypeScript SDK and mistralai Python SDK.

Prerequisites

  • Completed mistral-install-auth setup
  • Valid MISTRAL_API_KEY environment variable set
  • Node.js 18+ or Python 3.9+

Instructions

Step 1: Basic Chat Completion

TypeScript (hello-mistral.ts)

import { Mistral } from '@mistralai/mistralai';

const client = new Mistral({ apiKey: process.env.MISTRAL_API_KEY });

async function main() {
  const response = await client.chat.complete({
    model: 'mistral-small-latest',
    messages: [
      { role: 'user', content: 'Say "Hello, World!" in a creative way.' },
    ],
  });

  console.log(response.choices?.[0]?.message?.content);
  console.log('Tokens used:', response.usage);
}

main().catch(console.error);

Python (hello_mistral.py)

import os
from mistralai import Mistral

client = Mistral(api_key=os.environ["MISTRAL_API_KEY"])

response = client.chat.complete(
    model="mistral-small-latest",
    messages=[
        {"role": "user", "content": "Say 'Hello, World!' in a creative way."}
    ],
)

print(response.choices[0].message.content)
print(f"Tokens: {response.usage}")

Step 2: Run the Example

# TypeScript
npx tsx hello-mistral.ts

# Python
python hello_mistral.py

Step 3: Streaming Response

Streaming delivers the first token in ~200ms instead of waiting 1-2s for the full response.

TypeScript

import { Mistral } from '@mistralai/mistralai';

const client = new Mistral({ apiKey: process.env.MISTRAL_API_KEY });

async function streamChat() {
  const stream = await client.chat.stream({
    model: 'mistral-small-latest',
    messages: [
      { role: 'user', content: 'Tell me a short story about AI.' },
    ],
  });

  for await (const event of stream) {
    const content = event.data?.choices?.[0]?.delta?.content;
    if (content) process.stdout.write(content);
  }
  console.log(); // newline
}

streamChat().catch(console.error);

Python

stream = client.chat.stream(
    model="mistral-small-latest",
    messages=[{"role": "user", "content": "Tell me a short story about AI."}],
)

for event in stream:
    content = event.data.choices[0].delta.content
    if content:
        print(content, end="", flush=True)
print()

Step 4: Multi-Turn Conversation

const messages: Array<{ role: 'system' | 'user' | 'assistant'; content: string }> = [
  { role: 'system', content: 'You are a helpful coding assistant.' },
  { role: 'user', content: 'What is the capital of France?' },
];

const r1 = await client.chat.complete({
  model: 'mistral-small-latest', messages,
});
const answer = r1.choices?.[0]?.message?.content ?? '';
console.log('A1:', answer);

// Continue the conversation
messages.push({ role: 'assistant', content: answer });
messages.push({ role: 'user', content: 'What about Germany?' });

const r2 = await client.chat.complete({
  model: 'mistral-small-latest', messages,
});
console.log('A2:', r2.choices?.[0]?.message?.content);

Step 5: JSON Mode (Structured Output)

const response = await client.chat.complete({
  model: 'mistral-small-latest',
  messages: [
    { role: 'user', content: 'List 3 programming languages with their year of creation as JSON.' },
  ],
  responseFormat: { type: 'json_object' },
});

const data = JSON.parse(response.choices?.[0]?.message?.content ?? '{}');
console.log(data);

Step 6: With Temperature and Token Limits

const response = await client.chat.complete({
  model: 'mistral-small-latest',
  messages: [{ role: 'user', content: 'Write a haiku about coding.' }],
  temperature: 0.7,   // 0-1, higher = more creative
  maxTokens: 100,     // cap output length
  topP: 0.9,          // nucleus sampling
});

Output

  • Working code file with Mistral client initialization
  • Successful API response with generated text
  • Console output showing response and token usage

Error Handling

ErrorCauseSolution
Import ErrorSDK not installedRun npm install @mistralai/mistralai
401 UnauthorizedInvalid API keyCheck MISTRAL_API_KEY is set
ERR_REQUIRE_ESMCommonJS projectUse import syntax or dynamic await import()
429 Rate LimitedToo many requestsWait and retry with backoff

Model Quick Reference

Model IDBest ForContext
mistral-small-latestFast, cost-effective tasks256k
mistral-large-latestComplex reasoning, analysis256k
codestral-latestCode generation, FIM256k
mistral-embedText/code embeddings8k
pixtral-large-latestVision + text (multimodal)128k

Resources

Next Steps

Proceed to mistral-core-workflow-a for production chat patterns or mistral-local-dev-loop for dev workflow setup.

Info
Name mistral-hello-world
Version v20260423
Size 5.69KB
Updated At 2026-04-28
Language