Skip to the page
Conch
DocsGitHub

Providers

LM Studio

Your LM Studio models, here

The models you already have in LM Studio, running right here — private, free and offline. Conch starts LM Studio’s server when it needs to.

  • Private
  • Free
  • Works offline
Connects with
A program on this computer
Your apps
Conch hands it their tools
To start
Free
  • NoYour files and commands
  • YesSaves memories itself
  • YesAsks before each step
  • YesWorks offline

Worth knowing first

  • A model runs at your computer’s speed; the first answer waits while LM Studio loads it.
  • Tool-capable models can use Conch’s files and connected apps. Commands need the OS sandbox and run without network access; chat-only models cannot take actions.

Connect it

  1. Open Settings → Providers and choose LM Studio.
  2. If LM Studio isn’t here, press Install LM Studio. Conch installs it with winget or Homebrew, or links to it on Linux.
  3. Download a model in LM Studio. It shows up in the picker by itself.

Conch finds LM Studio where it keeps itself, and its server on whatever port it last used. If the server is off, Conch starts it and leaves a note under Fixed on its own.

What you get

The models you already have in LM Studio, running on this computer: private, free and offline. Conch reads what each one can do (pictures, tools, thinking) from LM Studio itself, and leaves out embedding models.

Conch runs the conversation itself: it keeps the thread, hands the model the tools of your apps, and asks you before anything changes. Memory, skills and the browser all work with models that can call tools.

Good to know

  • A small model gets a lean setup. When the model reads little at once (LM Studio often loads one with a small window), Conch gives it short instructions and hands it tools as it needs them. Load the model with a bigger context length in LM Studio, and Conch uses it.
  • The first answer waits for the model. LM Studio loads a model when it’s first asked. Conch says so in the chat while it loads.
  • A key, only if you turned one on. If Require Authentication is on in LM Studio, create a token under Developer → Server Settings → Manage Tokens and paste it on LM Studio’s page.
  • It answers offline. When the internet drops, Conch can hand the chat to it. See Offline and at a limit.