Open source

The self-hosted frontend the power users all run

If you want total control and zero subscription, this is where the advanced users live.

Run large language models locally with one command

The de-facto local model runtime: one command to run open models, free and open-source.

Free character chat that runs locally or in the cloud

The most generous free tier in the category: unlimited chat on locally run models, 600 cloud messages a day, no card required.

Self-hosted AI platform with personas, RAG and a 500k+ community

The power-user pick: run any model locally or in the cloud with zero data retention and full control.

Open-source ChatGPT alternative that runs fully offline

A clean local assistant: open-source, offline-first, with an OpenAI-compatible API server built in.

Open-source chat UI with a community agent marketplace

A serious open-source option: self-host it, plug in any model, and browse community personas — more assistant than romantic partner.

On-device AI chat with your documents, local or cloud models

Document-grounded private AI: chat with your files on desktop, connect local or hosted models.

The veteran local-model front end, now a no-install desktop app

Formerly text-generation-webui: not a companion app, but the plumbing most local companion setups are built on.

Open-source roleplay chat with community cards and no limits

The open-source flag behind the tavern scene: free web chat, community cards, plug in your own model.

Browser-based character chat with card importing

A lighter alternative to SillyTavern that runs in a browser tab with no install.

The original open-source tavern frontend for character roleplay

The grandfather of the tavern scene, now at version 2: free, open-source, plug in any model and any card.

Open-source multi-bot chat with persistent memory

Free and genuinely capable, with chat memory and multi-bot scenes.

Local story generation with no cloud dependency

The classic offline story engine: clunky, but it never phones home.