Mistral AI: Mistral Small 3.2 24B Instruct
mistral/mistral-small-24b-instruct-2506
Access Mistral Small 3.2 24B Instruct from Mistral AI using Puter.js AI API.
Get Started// npm install @heyputer/puter.js
import { puter } from '@heyputer/puter.js';
puter.ai.chat("Explain quantum computing in simple terms", {
model: "mistral/mistral-small-24b-instruct-2506"
}).then(response => {
document.body.innerHTML = response.message.content;
});
<html>
<body>
<script src="https://js.puter.com/v2/"></script>
<script>
puter.ai.chat("Explain quantum computing in simple terms", {
model: "mistral/mistral-small-24b-instruct-2506"
}).then(response => {
document.body.innerHTML = response.message.content;
});
</script>
</body>
</html>
# pip install openai
from openai import OpenAI
client = OpenAI(
base_url="https://api.puter.com/puterai/openai/v1/",
api_key="YOUR_PUTER_AUTH_TOKEN",
)
response = client.chat.completions.create(
model="mistral/mistral-small-24b-instruct-2506",
messages=[
{"role": "user", "content": "Explain quantum computing in simple terms"}
],
)
print(response.choices[0].message.content)
curl https://api.puter.com/puterai/openai/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_PUTER_AUTH_TOKEN" \
-d '{
"model": "mistral/mistral-small-24b-instruct-2506",
"messages": [
{"role": "user", "content": "Explain quantum computing in simple terms"}
]
}'
Model Card
Mistral Small 3.2 24B Instruct is a 24-billion-parameter chat model from Mistral AI, released in June 2025 as an update to Mistral Small 3.1.
It keeps the same 24B architecture and 128K context window as 3.1 but improves instruction following, from 82.75% to 84.78% on Mistral's internal accuracy benchmark, and cuts repetitive or infinite generations from 2.11% to 1.29%. Function calling is more robust, and Wildbench v2 and Arena Hard v2 scores rose to 65.33% and 43.1%.
On coding it scores 92.90% on HumanEval Plus and 78.33% on MBPP Plus (pass@5). It accepts text and image inputs, scoring 94.86% on DocVQA and 87.4% on ChartQA for document and chart understanding.
It is released under Apache 2.0, making it a fit for chat, tool-calling agents, and document or image understanding tasks where open licensing and low latency matter more than raw model size.
Context Window 128K
tokens
Max Output 128K
tokens
Input Cost $0.08
per million tokens
Output Cost $0.2
per million tokens
Release Date N/A
Model Playground
Try Mistral Small 3.2 24B Instruct instantly in your browser.
This playground uses the Puter.js AI API — no API keys or setup required.
More AI Models From Mistral AI
Find other Mistral AI models →
Mistral Small Creative
Mistral Small Creative is an experimental model from Mistral AI's Labs line, released in December 2025 and positioned for creative writing rather than general-purpose use. It targets creative writing, narrative generation, roleplay and character-driven dialogue, alongside general instruction following and conversational agents. It runs on a 32K context window and supports function calling, structured outputs, and document question answering. Mistral has not published benchmark scores for this model. On Mistral's own platform it has already been marked deprecated with a March 2026 sunset date, with Ministral 3 8B recommended as the replacement. For developers, it is best suited to fiction, dialogue, and other imaginative writing tasks rather than coding, reasoning, or long-context work, where Mistral's general-purpose Small or Medium models are a better fit.
ChatMixtral 8x7B Instruct v0.1
Mixtral 8x7B Instruct v0.1 is a sparse mixture-of-experts chat model from Mistral AI, released in December 2023 as the instruction-tuned version of the base Mixtral 8x7B model. It combines 8 feedforward expert blocks of 7B parameters each into one network, for 46.7 billion total parameters, of which a router activates about 12.9 billion per token. This gives it response quality closer to a much larger dense model while running with roughly 6x faster inference than Llama 2 70B. Mistral reports it matches or outperforms GPT-3.5 on most standard benchmarks, scoring 8.30 on MT-Bench versus GPT-3.5's 8.32, and outperforms Llama 2 70B on most benchmarks tested, including a 70.6% MMLU score. It handles English, French, German, Spanish, and Italian. With a 32K context window and Apache 2.0 licensing, it's a low-cost option for general chat and multilingual tasks that don't need a full 70B-class dense model.
Frequently Asked Questions
You can access Mistral Small 3.2 24B Instruct by Mistral AI through Puter.js AI API. Include the library in your web app or Node.js project and start making calls with just a few lines of JavaScript — no backend and no configuration required. You can also use it with Python or cURL via Puter's OpenAI-compatible API.
Yes, it is free if you're using it through Puter.js. With the User-Pays Model, you can add Mistral Small 3.2 24B Instruct to your app at no cost — your users pay for their own AI usage directly, making it completely free for you as a developer.
| Price per 1M tokens | |
|---|---|
| Input | $0.08 |
| Output | $0.2 |
Mistral Small 3.2 24B Instruct supports a context window of 128K tokens. For reference, that is roughly equivalent to 256 pages of text.
Mistral Small 3.2 24B Instruct can generate up to 128K tokens in a single response.
Yes — the Mistral Small 3.2 24B Instruct API works with any JavaScript framework, Node.js, or plain HTML through Puter.js. Just include the library and start building. See the documentation for more details.
Get started with Puter.js
Add Mistral Small 3.2 24B Instruct to your app without worrying about API keys or setup.
Read the Docs View Tutorials