Mistral Medium 3.5
Neural Network
Mistral Medium 3.5 is a Mistral AI model for code, agentic scenarios, and complex tasks with text and image input support.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 262 144 tokens
- Max answer length
- 209 715 tokens
- Reasoning effort
- none, high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Mistral Medium 3.5?
Mistral Medium 3.5 is a dense 128-billion parameter model from the Mistral AI team. It accepts text and images as input, outputs text, and follows instructions precisely, making it reliable for long chains of thought in agentic scenarios, coding, and complex tasks where reasoning is key. Via BotHub, the model is accessible from Russia without a VPN or foreign card, with pay-as-you-go billing in rubles for tokens used, no subscription required. Over 250 models are available in one window: switch between them while working and compare responses. A unified OpenAI-compatible API allows switching models without rewriting integrations. Requests are encrypted, and for companies, we offer contracts, invoices, EDI, and an admin panel with limits. It is useful for analyzing unfamiliar repositories and explaining their logic, writing and debugging functions, building agents for multi-step tasks, extracting meaning from interface screenshots or diagrams, and preparing technical project documentation.Frequently asked questions about Mistral Medium 3.5
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 262,144 token context. It is suitable for working with long documents and code, analytics, refactoring, debugging, and preparing structured responses. It accepts images and documents as input, can reason before answering, and call external functions.
Up to 209,715 tokens in a single response — enough for a multi-page report, restructuring a large document, or generating an entire code module. The actual length depends on the task and your prompt: usually, the model uses far less than the full available volume.
Yes, reasoning before answering is confirmed: the model can break down a task into steps before providing a result. This helps with math, logic, code analysis, and multi-step instructions. The difference is less noticeable on simple queries, but on complex ones, the response is more precise.
Yes. The model supports function calling and structured JSON output, making it easy to integrate into agents, chatbots, and pipelines where the response goes directly into code. In BotHub, this works via a unified OpenAI-compatible API — you can switch models without rewriting the integration.
Images and documents — yes: send a screenshot, diagram, table, or file and ask to analyze the content. Audio and video input is not noted in our data. Image, audio, or video generation is also not noted for this model — BotHub has separate models for such tasks.
Languages are not specified in our data, so we won't promise anything specific. It's easier to test with your own task: send a typical query and compare the result with another model in the same window. Payment and access are in rubles, without a VPN or foreign card.