Forge / Local & Offline

GPT4All

Open source

The onboarding ramp to local LLMs. Less power-user than LM Studio, easier for a first install.

No cardMM-friendlySelf-hostableAPI

Free tier

MIT desktop chat. One-click model installer. Fully offline after download.

Limits

Weaker agent story. Model catalog lags LM Studio. Not for production serving.

Best for

  • First local install
  • Air-gapped chat
  • Non-developers

How to use it well

  1. Install, pick a recommended 7B, wait for the download, chat.
  2. Enable LocalDocs only on folders you accept being embedded on-disk.
  3. Graduate to LM Studio/Ollama when you need an OpenAI endpoint for agents.

Myanmar notes

Give this to a teammate who will not use a terminal. You, use Ollama or LM Studio.