Meowbert Docs

Appearance

Bring Your Own Provider & Subscriptions ​

Meowbert lets you power agent tasks using your own AI subscriptions or custom API keys via Bring Your Own (BYO) mode.

You can configure BYO settings anytime by navigating to User Settings → Subscription & Providers.

When BYO is disabled, tasks use the provider selected by your server admin. Admins manage these connections in Admin → AI providers: add a base URL and API key, select a provider to use it, or remove a saved connection. The first provider is selected automatically. Switching providers applies to new runs; runs already in progress keep their current connection. Removing the selected provider requires selecting another before new platform AI runs can start. Model choices remain under Admin → Model.

Supported BYO options ​

1. Sign in with ChatGPT (experimental) ​

If you already have a personal or team ChatGPT Plus, Pro, or Team subscription, you can connect your OpenAI account directly to Meowbert without creating or paying for separate API keys.

How it works ​

  1. Go to Subscription & Providers and select the Sign in with ChatGPT tab.
  2. Click Sign in with ChatGPT to open the activation modal.
  3. Meowbert will generate a unique one-time device code. Click Open OpenAI Activation or visit the activation page.
  4. Sign in to your ChatGPT account and confirm the code.
  5. Once authorized, Meowbert automatically activates ChatGPT BYO mode. You can optionally force a model for every ChatGPT BYO task.

Experimental

Sign in with ChatGPT isn't an official OpenAI integration. Meowbert uses the same sign-in flow as other open-source coding tools (the implementation follows opencode's), which reuses the Codex CLI's device-code login. OpenAI can change or block it at any time, so it may stop working without notice. Check that this use is allowed under your OpenAI account's terms.

For a more reliable setup, run a subscription proxy such as CLIProxyAPI and connect it as a Custom API Key instead.

Managing your connection ​

  • Force default model: Enter an exact model identifier if every ChatGPT BYO task should use the same model. Leave it blank to use the model selected for each task.
  • Disconnect anytime: Click the Disconnect button to unlink your ChatGPT account and restore standard workspace billing.

2. Custom API Key (OpenAI-compatible) ​

If you prefer using an API key from OpenAI, OpenRouter, DeepSeek, or a self-hosted inference server (such as vLLM or Ollama), you can configure a custom endpoint:

  1. Select the Custom API Key tab.
  2. Enter your Base URL (e.g. https://api.openai.com/v1 or https://openrouter.ai/api/v1).
  3. Enter your Model identifier (e.g. gpt-4o, deepseek-chat).
  4. Paste your API key.
  5. Click Fetch Models to verify the connection and discover available models, then click Use Custom API Key.

Switching or disabling BYO ​

  • You can switch between ChatGPT Account and Custom API Key at any time.
  • To pause using your own provider and go back to the server's provider and any usage limits your admin has set, click Disable BYO in the top right of the section.