Mistral Large 4
Neural Network
Mistral Large 4 is a multimodal Mistral AI model with 512K context: Mistral Large 4 API for reasoning, coding, and agents.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 1 048 576 tokens
- Max answer length
- 262 144 tokens
- Reasoning effort
- none, high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
import OpenAI from 'openai';
const openai = new OpenAI({
apiKey: '<your bothub access token>',
baseURL: 'https://openai.bothub.chat/v1'
});
// Sync - Text generation
async function main() {
const chatCompletion = await openai.chat.completions.create({
messages: [{ role: 'user', content: 'Say this is a test' }],
model: 'mistral-large-4-0',
});
}
// Async - Text generation
async function main() {
const stream = await openai.chat.completions.create({
messages: [{ role: 'user', content: 'Say this is a test' }],
model: 'mistral-large-4-0',
stream: true
});
for await (const chunk of stream) {
const part: string | null = chunk.choices[0].delta?.content ?? null;
}
}
main();
How it works Mistral Large 4?
Mistral Large 4 is Mistral AI's flagship model that accepts text and images, designed for reasoning, coding, and agentic scenarios. The 512K token context allows you to process large codebases or batches of documents in a single request, while the multimodal input can analyze attached diagrams or interface screenshots. Through BotHub, the model is available without VPN or foreign cards: pay in rubles with Russian cards per token, no subscription, unused tokens do not expire. Access over 250 neural networks in one window, switch between them without rewriting integrations using a unified OpenAI-compatible API. Chats are encrypted via AES-GCM. For companies, we offer contracts, invoices, electronic document management, and an admin panel with limits. Use cases: refactor and debug large projects, build agents with function calling and external tools, summarize long reports with chart screenshots, or solve STEM problems step-by-step.Frequently asked questions about Mistral Large 4
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
This is a Mistral text model with a reasoning mode, accepting images and documents as input. It is suitable for working with code: writing, refactoring, and debugging, as well as analyzing long reports and contracts, analytics, STEM tasks, and building agents with function calling.
The output limit is 262,144 tokens, and the context window is 524,288 tokens. This is enough to generate extensive documentation, a large code module, or a detailed analysis of a large document in a single pass without splitting the task.
Yes, the reasoning mode is confirmed. The model breaks down the task step-by-step before providing the final answer, which helps with math, logic, data analysis, and debugging code. In BotHub, you can compare its output with other models in one window.
Yes, both function calling and structured JSON output are supported. This allows you to build agents, connect external services and databases, and receive predictable schemas for parsing. With BotHub's unified OpenAI-compatible API, you won't need to rewrite your integration.
The model accepts images and documents as input: you can send a diagram, interface screenshot, table, or PDF and ask it to analyze the content. We have no information regarding audio or video, so please expect support for text, images, and files.
Mistral models traditionally work with multiple languages, and Russian queries are usually handled easily by this version. The easiest way to check is in practice: BotHub provides access without VPN or foreign cards, with payment in rubles per token.