Get Started

Chat with
any model

A beautiful, open-source interface for running and interacting with AI models.

Start Chatting

Run local models with Ollama and manage your AI workspace with ease.

Learn more

Flexible, extensible, and ready to be part of your workflow.

View on GitHub 
$ git clone https://github.com/eissapk/tokn
$ cd tokn && npm install
$ npm run dev

Features

A fast, focused chat window for the models on your machine, with the tools you reach for every day.

Private by default

Models run on your own machine through Ollama. Your prompts and chats never leave it.

Thinking mode

Switch on thinking and watch reasoning models work through hard problems step by step.

Attach files

Add files to a message and ask questions about them, right in the chat.

Search every chat

Chats are saved as you go. Star the ones that matter and find anything with Ctrl+K.

Context meter

See how much of the model's memory a conversation is using before it starts to forget.

Code that copies

Answers render as clean markdown, with a copy button on every code block.

Get started

No accounts and no API keys. If Ollama runs on your computer, so does TOKN.

  1. Install Ollama

    Download it from ollama.com and start its server.

    $ ollama serve

  2. Pull a model

    Grab any model from the Ollama library. Llama 3.2 is a good first pick.

    $ ollama pull llama3.2

  3. Start chatting

    Open TOKN, pick the model from the menu at the top, and say hello.

    Open the chat

Made for the keyboard

Start, find and switch between chats without reaching for the mouse.

  • New chat Ctrl N
  • Search chats Ctrl K
  • Previous / next chat Alt ↑ / ↓

Open TOKN and start a conversation with any model on your machine.

Start Chatting
View on GitHub