Dolphin Mistral 24B Venice Edition
Neural Network
Dolphin Mistral 24B Venice Edition is a fine-tuned version of Mistral Small 24B Instruct by dphn.ai for text tasks.
Max answer length
(in tokens)
Context size
(in tokens)
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Dolphin Mistral 24B Venice Edition?
Dolphin Mistral 24B Venice Edition is a fine-tuned version of Mistral Small 24B Instruct 2501, prepared by the dphn.ai team together with Venice.ai. It inherits the architecture and 24 billion parameters from the base Mistral model, but differs in its fine-tuning layer: the model is trained to follow user instructions precisely. Via BotHub, it is available without a VPN or foreign card, and you pay in rubles only for the tokens used—no subscriptions, and the balance does not expire. With over 250 neural networks in one window, you can easily run the same prompt through multiple models and compare answers. Developers have access to a unified OpenAI-compatible API, chats are encrypted via AES-GCM, and BotHub provides contracts, invoices, and electronic document management for companies, along with an admin panel and limits. The model is useful for drafting text, summarizing long documents, breaking down correspondence, building an assistant via API, or testing how different fine-tuned Mistral versions handle a task.Frequently asked questions about Dolphin Mistral 24B Venice Edition
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 text model from cognitivecomputations: it works with text, accepts documents and images. The 128,000-token context allows you to analyze long files, write and edit code, prepare drafts, summarize correspondence, and extract data from scans and screenshots.
The model outputs up to 8,192 tokens per response, which is roughly several pages of text. If you need extensive material, break the task into parts: ask it to continue from a specific point or generate by sections, then assemble the result.
A separate reasoning mode is not noted in our data, but that doesn't mean it doesn't exist. Like any text model, you can ask it to write out the solution step-by-step—this technique usually significantly improves answers to logic and math problems.
Support for function calling and strict JSON mode is not noted in our data, so we cannot guarantee it. Usually, an explicit instruction in the prompt helps: describe the required schema and ask it to return only JSON, then verify the result with a validator on your side.
Images and documents—yes, the model accepts them as input along with text: you can show a screenshot, diagram, or file and ask it to analyze the content. Audio and video input is not noted in our data; BotHub has separate models for such tasks.
We do not have data on the languages the model works with, so we won't promise anything specific. Test it with your task: in BotHub, you pay per token, access is without VPN or foreign cards, and you can switch to another model in the same window.