Mistral Medium 3
Neural Network
Mistral Medium 3 is a corporate language model from Mistral AI for reasoning and multimodal tasks.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 131 072 tokens
- Max answer length
- 104 857 tokens
- Capabilities
- function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Mistral Medium 3?
Mistral Medium 3 is a model from the French company Mistral AI, created for corporate scenarios: it combines advanced reasoning and multimodal capabilities with an operating cost approximately eight times lower than comparable solutions. In BotHub, it is available from Russia without a VPN or foreign card: you pay for tokens with Russian cards as you go, without a subscription, and unused tokens do not expire. Nearby, in one window, more than 250 neural networks are open, so it is easy to compare answers and switch to another model without rewriting the integration: connection is via a unified OpenAI-compatible API, correspondence is encrypted, and companies have access to contracts, invoices, electronic document management (EDO), and an admin panel with limits. It is useful when you need to analyze a voluminous document and summarize it, build analytics and explain the reasoning process, process an image together with a text query, build an internal assistant for the team, or use the model as the basis for a support chatbot.Frequently asked questions about Mistral Medium 3
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 universal text model from Mistral with a context of 131,072 tokens. It is suitable for code work and refactoring, analyzing long documents, analytics, and STEM tasks, as well as for chatbots and assistants with tool calling. Additionally, it accepts images and files as input.
The output limit is up to 104,857 tokens per response, which is the volume of a full-fledged document or a large code module. Together with the 131,072-token context window, you can upload large material and get a detailed analysis without splitting it into parts.
A separate extended reasoning mode is not noted in our data. However, you can ask the model to write out the solution step-by-step directly in the prompt — for logical and STEM tasks, this usually improves quality. If you need reasoning models, switch to them in the same BotHub window.
Yes, the model supports function calling and structured JSON output. This allows you to connect external tools, databases, and services, build agents, and receive predictable answers according to a schema. It works via the unified OpenAI-compatible BotHub API, so the integration does not need to be rewritten when changing models.
The model accepts text, images, and documents as input, including native file processing. You can send a screenshot, diagram, PDF, or table and get an analysis, data extraction, or summary. Audio and video support is not noted in our data.
Mistral Medium 3 is a multilingual model, and it works with Russian-language queries: it answers, summarizes documents, and helps with code and phrasing. It is easy to check the quality on your own tasks — in BotHub, you only pay for the tokens actually consumed and can compare the answer with other models.