MegaBrain University · guide

Kilo Code with a Custom OpenAI-Compatible Provider: Claude, GPT-6 and Open Models

Updated November 11, 2026 · 10 min setup

What is Kilo Code

Kilo Code is an open-source AI coding agent for VS Code and JetBrains IDEs, with a terminal version, the Kilo CLI (a fork of OpenCode). It plans, edits files, runs commands, uses MCP servers and switches between agent modes.

Kilo can talk to any OpenAI-compatible API through a custom provider. Connecting MegaBrain gives it every major model behind one key and one balance, so you can put a frontier model on architecture work and a cheap one on routine edits.

Setup, step by step

1. Step 1. Get a MegaBrain API key

   Sign up with GitHub, Google or email, add funds on the Balance page and create a key on the API key page. The key works for every text, image and video model, and each request is paid from your balance — no subscription.

   Put the key in an environment variable so it never lands in a config file you share:

   # ~/.zshrc or ~/.bashrc
   export MEGABRAIN_API_KEY="your-megabrain-key"

   Copy

   curl https://getmegabrain.com/api/gateway/v1/chat/completions \
     -H "Authorization: Bearer $MEGABRAIN_API_KEY" \
     -H "Content-Type: application/json" \
     -d '{"model": "anthropic/claude-sonnet-5.5", "messages": [{"role": "user", "content": "Say OK"}]}'

   Copy

   Check the key before you touch the agent. A JSON reply with choices means the key, the balance and the model all work:

2. Step 2. Install Kilo Code

   In VS Code or a JetBrains IDE, install Kilo Code from the extension marketplace. For the terminal: npm install -g @kilocode/cli (the command is kilo).

3. Step 3. Add MegaBrain in the extension

   Open Kilo’s settings (gear icon) → Providers → Custom provider and fill in:

   Provider ID:   megabrain
   Display name:  MegaBrain
   Provider API:  OpenAI Compatible
   Base URL:      https://getmegabrain.com/api/gateway/v1
   API key:       your-megabrain-key
   Models:        Fetch models (or type anthropic/claude-sonnet-5.5)

   Copy

   Click Submit, then choose MegaBrain → anthropic/claude-sonnet-5.5 in the model picker.

4. Step 4. Or configure the Kilo CLI

   The CLI reads ~/.config/kilo/kilo.jsonc. Keep the key in the global file: Kilo ignores {env:…} in project config.

   {
     "$schema": "https://app.kilo.ai/config.json",
     "model": "megabrain/anthropic/claude-sonnet-5.5",
     "provider": {
       "megabrain": {
         "npm": "@ai-sdk/openai-compatible",
         "name": "MegaBrain",
         "options": {
           "baseURL": "https://getmegabrain.com/api/gateway/v1",
           "apiKey": "{env:MEGABRAIN_API_KEY}"
         },
         "models": {
           "anthropic/claude-sonnet-5.5": {
             "name": "Claude Sonnet 5.5",
             "tool_call": true,
             "limit": { "context": 1000000, "output": 32000 }
           },
           "moonshotai/kimi-k3.1-pro": {
             "name": "Kimi K3.1 Pro",
             "tool_call": true,
             "limit": { "context": 1048576, "output": 32000 }
           }
         }
       }
     }
   }

   Copy

   Set tool_call: true (without it, file edits and the terminal tool do not work) and limit.context (without it, Kilo never compacts a long conversation). Run kilo, then /models to switch.

Which model to choose

Claude Sonnet 5.5 is the default for coding in Kilo. Use Opus 5.5 or GPT-6.1 Sol for architecture and hard bugs, Kimi K3.1 Pro or GLM 5.3 for everyday edits, DeepSeek V4.1 Flash for bulk mechanical changes.

Every model with its price and context window is in the catalog. They all work with the same key.

What you can do

Business use cases

Plan the migration of this Express 4 app to Express 5: list breaking changes that affect us, with file paths. Then, file by file, apply the changes and run the tests after each file.

Troubleshooting

FAQ