pi-openai-compat

@billjr99/pi-openai-compat is an extension for the pi coding agent terminal harness that registers OpenAI-compatible LLM endpoints as native providers. Once installed and configured, models from external providers appear directly in pi’s /model list and Ctrl+L picker alongside the built-in Anthropic, OpenAI, and Google models, with no custom model selection UI required. The extension uses pi’s registerProvider API, the same mechanism pi uses internally for its first-party integrations.

Multiple providers can be active simultaneously, and their models appear together in /model under their respective provider labels. On deregistration, pi’s built-in unregisterProvider restores the original model list automatically. Model lists are cached in a local config.json at startup, so no network call is required to re-register providers when pi restarts. API keys are stored in ~/.config/pi-openai-compat/config.json.

The extension is published to npm as a scoped package and is also installable directly from GitHub.

The extension is hosted on GitHub and published to npm:

Installation

# From npm
pi install npm:@billjr99/pi-openai-compat

# From GitHub
pi install git:github.com/BillJr99/pi-openai-compat

If pi is already running when you install, type /reload first.

Supported Providers

Commands

The extension provides two commands:

• /compat-login: Walks through a short wizard to select a provider, confirm the base URL, and enter an API key. The provider’s models appear in pi’s /model command immediately, and the command can be run again to add additional providers.
• /compat-logout: Unregisters a provider. If multiple providers are registered, the user is prompted to select which one to remove; when the last provider is removed, the previously active built-in model is restored.

Typical Workflow

/compat-login
  → pick Ollama
  → press Enter to accept http://localhost:11434/v1
  → models added to /model

/model
  → select desired model

/compat-logout
  → previous model restored automatically