TMA-1/Dolphin-Mistral-24B-Venice-Edition

VISIONPricing:Input $0.867 / Output $1.61Concurrent Unit Cost:2Model Size:24BQuant:FP8Context Size:32kTool Calling:SupportedPublished:Sep 10, 2026License:apache-2.0Architecture:Transformer Open Weights Featherless Exclusive Cold

TMA-1/Dolphin-Mistral-24B-Venice-Edition is a 24 billion parameter Mistral-based language model developed collaboratively by Dolphin and Venice.ai, with a context length of 32768 tokens. It is specifically fine-tuned to be an uncensored, steerable general-purpose model, giving users full control over system prompts and alignment. This model is designed for applications requiring custom ethical guidelines and data privacy, contrasting with more restrictive commercial LLMs.

Loading preview...

What is Dolphin Mistral 24B Venice Edition?

Dolphin Mistral 24B Venice Edition is a 24 billion parameter model, a collaborative effort between Dolphin and Venice.ai, built upon the Mistral architecture. Its primary goal is to provide an uncensored and highly steerable general-purpose language model, offering users complete control over its behavior and alignment.

Key Capabilities & Differentiators

  • Uncensored by Design: Explicitly trained to follow instructions without ethical or safety reservations, allowing users to define its guidelines.
  • User Control: Unlike commercial models, it grants full control over system prompts and alignment, preventing unexpected changes or deprecations.
  • Data Privacy: Designed to keep user queries private, addressing concerns about data usage by third-party LLM providers.
  • General Purpose: Aims to function as a versatile model, similar to leading commercial LLMs, but with enhanced user autonomy.
  • Mistral Chat Template: Maintains Mistral's default chat template for ease of integration.

Ideal Use Cases

  • Custom AI Applications: Businesses and developers needing an LLM where they dictate the ethical boundaries and response style.
  • Research & Development: Exploring the capabilities of uncensored models and custom alignment strategies.
  • Private Deployments: Scenarios where data privacy and control over model behavior are paramount.
  • Flexible System Prompts: Environments where dynamic and highly specific system prompts are crucial for application functionality.

Finetunes
1 models