pi-setup-custom-providers

Interactive wizard to manage custom LLM providers and models for pi coding agent

Packages

Package details

extension

Install pi-setup-custom-providers from npm and Pi will load the resources declared by the package manifest.

$ pi install npm:pi-setup-custom-providers
Package
pi-setup-custom-providers
Version
1.0.1
Published
May 3, 2026
Downloads
335/mo · 36/wk
Author
davidyusaku
License
MIT
Types
extension
Size
187.5 KB
Dependencies
2 dependencies · 0 peers
Pi manifest JSON
{
  "extensions": [
    "./src/index.ts"
  ]
}

Security note

Pi packages can execute code and influence agent behavior. Review the source before installing third-party packages.

README

pi-setup-custom-providers

Interactive wizard to manage custom LLM providers and models for the pi coding agent.

Features

  • Provider Management — Add, edit, remove custom providers (Ollama, OpenRouter, Groq, Together, DeepSeek, Mistral, Anthropic, Google, llama.cpp, LM Studio, enowX Labs, CLI Proxy API, and more)
  • Model Discovery — Auto-discover models from provider APIs with rich metadata (context window, max tokens, reasoning support, image input, pricing)
  • Health Checks — Background connectivity checks with per-provider status
  • Chat Testing — Test chat completions against any configured model
  • Compat Auto-Detection — Automatically applies compatibility settings for known providers
  • Thinking Level Maps — Configure per-model reasoning effort mappings (enables xhigh)
  • Import/Export — Share provider configurations as JSON
  • Ollama Cloud Tier Detection — Probes which models are free vs pro subscription
  • Batch Operations — Multi-select models for bulk edits (limits, reasoning, image support)

Install

pi install npm:pi-setup-custom-providers

Or from GitHub:

pi install https://github.com/davidyusaku-13/pi-setup-custom-providers

Usage

In pi, run:

/setup-custom-providers

This opens an interactive TUI wizard where you can:

  1. Add providers from presets or custom URLs
  2. Discover models from the provider's API
  3. Configure model capabilities (reasoning, image input, context window, etc.)
  4. Test connectivity and chat completions
  5. Manage compatibility settings

Supported Providers

Provider Type Auto-Discovery
Ollama (local) Local ✅ via /api/tags + /api/show
Ollama Cloud Cloud ✅ with tier probing
llama.cpp Local ✅ via /v1/models + /props
LM Studio Local ✅ via /v1/models
OpenRouter Cloud ✅ with pricing
Together AI Cloud
Groq Cloud
DeepSeek Cloud
Mistral Cloud
Anthropic Cloud
Google Gemini Cloud
enowX Labs Gateway ✅ with tiers
CLI Proxy API Proxy

License

MIT