🐙 Nodink — Local LLM Studio Run open-source language models directly on your machine. No subscription, no data leaving your device. Nodink gives you a full studio-like experience to chat, experiment, and build with AI — completely free and open.
🤔 Why Nodink? Ollama is powerful but painful — no real interface, everything through the terminal, not built for people who just want to use a model. LM Studio gets it almost right, but stops at your own machine. Nodink started from a simple frustration : what if you could have the studio experience, but tap into other people's GPUs when yours isn't enough? That's the idea. Simple interface, full control, and a peer-to-peer GPU network coming when the studio is solid.
✨ Features
🖥️ Local first
Load and run open LLMs locally with full control over the model, quantization level, and context window. Your hardware, your rules.
🔑 Commercial models support
Connect any commercial AI provider — OpenAI, Anthropic, Mistral and more — by simply dropping in your API key. Switch between local and cloud models instantly.
🧠 Context & Agents
Define your context, set system prompts, and build lightweight agents powered by open LLMs. No platform lock-in.
⚖️ Response comparison
Run the same prompt across multiple models side by side and compare outputs. Find what works best for your use case.
⚙️ Full model control
Choose your model, adjust quantization, tune generation parameters — temperature, top-p, repetition penalty — directly from the interface.
🛠️ Tech Stack
AI & Backend — Python · Hugging Face Transformers · PyTorch · Flask Frontend — HTML · CSS · JavaScript Models — Qwen3-0.6B · Qwen3-4B (and more to come)
🌊 Coming soon — Depthlnk
A peer-to-peer GPU network built into Nodink.
Use credits to run inference on other users' hardware when your machine isn't enough Become a worker — share your GPU, earn credits No central server. Just nodes.
📄 License
This project is licensed under the PolyForm Noncommercial License 1.0.0. Free for personal and non-commercial use. Commercial use requires a separate agreement.
Built from home, one commit at a time.