xiry.

AI · AI model

Dolphin 3.0

Uncensored, open-source fine-tuned LLM designed for local execution without guardrails

From the Xiry library · Updated 3 October 2026

Visit huggingface.co ↗

Dolphin 3.0 is a series of open-weight large language models fine-tuned by Eric Hartford, based on architectures such as Llama and Qwen. It strips away standard corporate alignment and censorship filters, making it compliant with any instruction. Users typically run it locally via Ollama or LM Studio for privacy and unrestricted assistance.

What it's for

  • Run an unfiltered AI assistant completely offline using Ollama or LM Studio
  • Generate creative writing or fiction without triggering safety refusals
  • Conduct security and red-teaming research without corporate guardrail interference
  • Fine-tune bespoke local agents on sensitive or specialized tasks

Features

  • Uncensored dataset training removing standard alignment refusals
  • Compatibility with local inference engines like Ollama, LM Studio, and llama.cpp
  • Built across multiple base architectures (Llama 3.1, Llama 3.2, Qwen 2.5)
  • Available in quantized formats for low-resource local hardware
  • Open dataset and weight availability on Hugging Face

Pricing

Open source. Free and open source

Details

Type
AI model
Category
AI
Topic
Large Language Models
Open source
Yes
Platforms
Linux, macOS, Windows, Self-hosted, Command line
For
Developers, AI researchers, and hobbyists wanting an unfiltered local LLM
Alternative to
ChatGPT, Claude, Meta Llama
Tags
uncensored, llm, local ai, ollama, open source, llama, qwen

Why it was saved

  • @okaashish, 11 February 2025: It removes corporate alignment restrictions and censorship, allowing developers to run an unfiltered assistant locally via Ollama or LM Studio without cloud subscriptions.

Compare

Related

  • Pinokio

    Browser that installs and runs open-source AI apps with one click

  • Hermes 3

    Open-weights fine-tune of Llama 3.1 built for agentic function calling.

  • AnythingLLM

    Private, on-device AI assistant to chat with documents and local models.

  • Stop Paying the OpenAI Tax: The Emerging Open-Source AI Stack

    Architectural guide breaking down developer-friendly open-source alternatives across modern AI layers

  • Local Deep Researcher

    Fully local web research and report writing assistant using Ollama or LMStudio

  • AgenticSeek

    Open-source, fully local autonomous AI agent that browses and writes code

Updated