Skip to main content
Kilo Code is an open-source coding agent for VS Code and JetBrains. It connects to Lyceum through its OpenAI Compatible provider.

Install

  1. Open the VS Code Extensions panel.
  2. Search for Kilo Code and click Install.
  3. Restart VS Code.

Configure

1

Open the provider settings

Click the Kilo Code icon in the VS Code side bar, open the settings with the gear icon, and go to the Providers tab.
2

Add a custom provider

Scroll to the bottom, click Custom provider, and fill in:
3

Pick the models

Under Models, Kilo lists every model from our /models endpoint, which confirms that the key and URL work. The list includes image and video models, so pick the chat models you want, then click Submit.
Kilo saves the provider to ~/.config/kilo/kilo.jsonc. The Kilo CLI (npm install -g @kilocode/cli) reads the same file.

Modes

Kilo Code has several modes, and you can point each at a different model: Kilo’s own search tool is off for custom providers, so a Lyceum model reads pages with webfetch when you give it a URL. To let it search, add web_search_options to the model’s options in ~/.config/kilo/kilo.jsonc. Lyceum runs the searches and the model answers with its sources:
~/.config/kilo/kilo.jsonc
Set it on each model, not on the provider. The model cites results as [1] and [2], and Kilo doesn’t show the source links, so ask for the URLs in your prompt. Give any search tool you add yourself a name other than web_search. See Web search for options and pricing.

Notes

  • Token limits. Kilo doesn’t read token limits from a custom provider. Set them per model in kilo.jsonc, for example "limit": { "context": 200000, "output": 131072 }, with numbers from the models page.

Troubleshooting