Kilo Code

Kilo Code connects through a custom provider that uses the OpenAI Compatible API.

Setup

1. In VS Code, open Kilo Code Settings (gear) and go to the Providers tab.
2. Scroll to the bottom and click Custom provider.
3. Fill in the dialog:

Field	Value
Provider ID	standardcompute
Display name	Standard Compute
Provider API	OpenAI Compatible
Base URL	https://api.stdcmpt.com/v1
API key	Your key from the dashboard
Models	StandardCompute (the router), picked from the fetched list

4. Click Submit. The model appears in Kilo's model picker.

Give each model a context limit (1000000 for the router) in kilo.json: Kilo uses it to size its context gauge and decide when to compact, and a missing limit makes it compact far too early.

Kilo CLI

Add the provider to ~/.config/kilo/kilo.json and export STANDARDCOMPUTE_API_KEY:

{
  "provider": {
    "standardcompute": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Standard Compute",
      "env": ["STANDARDCOMPUTE_API_KEY"],
      "options": { "baseURL": "https://api.stdcmpt.com/v1" },
      "models": {
        "StandardCompute": { "name": "Standard Compute router", "limit": { "context": 1000000, "output": 32000 } }
      }
    }
  },
  "model": "standardcompute/StandardCompute"
}

Check the connection

Start a new task, ask for a small file edit, and confirm the request appears in Request Logs.

Pinning a model

To use one specific model, add its exact API ID from the Models page (for example moonshotai/kimi-k3) and set its context limit to that model's advertised window.

note

Your usage is shown in the Standard Compute dashboard. Kilo's local price estimate may differ.