Run Kilo Code on Mingles Router
Point the Kilo Code extension at Mingles Router with the built-in “OpenAI Compatible” provider — base URL, key, model id, done.
Pick your model and paste your key — every command and value on this page updates live.
Endpoint: https://router.mingles.ai/v1 · key stays in your browser ·
no key entered? commands show <your-key>.
Install + configure
1. Install the Kilo Code extension
From the VS Code Extensions panel search Kilo Code and install it, or run:
code --install-extension kilocode.Kilo-Code 2. Set the API Provider to “OpenAI Compatible”
Open the Kilo Code panel → settings (gear) → API Provider → OpenAI Compatible.
3. Fill Base URL, API Key, Model ID
Use the values in the table below. When the Base URL is valid, Kilo auto-fetches the model list from /v1/models — pick {{model}} or type it.
4. Start coding
Open the chat, choose a mode (Code / Architect / Ask) and go.
Exact values
| API Provider | OpenAI Compatible |
| Base URL | https://router.mingles.ai/v1 |
| API Key | <your-key> |
| Model ID | moonshotai/Kimi-K2.6 |
Already using Kilo Code? Add a profile
Create a new API configuration profile
Settings (gear) → API Configuration → add a profile → OpenAI Compatible, then fill Base URL, API Key and Model ID from the table above. Switch profiles per project without touching your other providers.
Switch model
Change the Model ID field to swap models on the same endpoint — no new key. Kilo Code is a fork of Cline, so Cline’s OpenAI-Compatible settings are identical.
Before you file a bug: read the limits
Reasoning models spend the output budget on internal thinking, output is capped at 8192 tokens, there is no KV cache, and there are no built-in web tools. Most “it broke on Mingles Router” reports are one of these. See Model limits & behavior →
Frequently asked
Which provider do I choose? +
“OpenAI Compatible”. Do not pick “OpenAI” — that one ignores a custom base URL.
It says “Model Not Found”. +
Make sure the Model ID is exactly moonshotai/Kimi-K2.6 (case-sensitive) and the Base URL ends in /v1. Kilo fetches the model list from the endpoint once the URL is valid.
Does this work in JetBrains and Cline too? +
Yes. The Kilo Code JetBrains plugin and Cline use the same “OpenAI Compatible” provider with the same three fields.