pi-params
Per-provider-per-model parameter overrides (temperature, presence_penalty, repetition_penalty, frequency_penalty, top_k, top_p, max_token)
6.2k listings for pi
691 results · page 12 of 15
Per-provider-per-model parameter overrides (temperature, presence_penalty, repetition_penalty, frequency_penalty, top_k, top_p, max_token)
Neuralwatt provider for pi — energy usage tracking and account quota in your status bar
Provider/model-scoped service_tier management extension for Pi
OMLX provider extension for pi - Access self-hosted Qwen models through a local MLX/OpenAI-compatible server
Token and cost statistics extension for pi coding agent sessions.
Pi provider extension for Ollama Cloud (ollama.com), direct connection without a local server.
Pi extension for managing custom model providers with discovery, metadata sources, generated config, and Pi-native TUI.
Provider-agnostic image generation for Pi coding agent
Universal complexity optimizer for all AI coding agents — Claude Code, Codex, Pi, Cursor, Windsurf, Copilot, Gemini CLI, Cline, Aider, OpenCode, and more.
Bonsai provider extension for pi-coding-agent
Headroom proxy extension for pi-coding-agent - routes requests through local Headroom compression proxy for 30-60% token savings, with CCR retrieval support
Usage indicator for pi — footer status bars and /usage command for Codex, Claude, Z.AI, Gemini, Antigravity, OpenCode Go, and Kiro
Production-ready /usage bars extension for pi (Codex, Claude, Z.AI, Gemini CLI, Antigravity)
Anthropic Claude models on Google Cloud Vertex AI / Gemini Enterprise Agent Platform for pi.
Use your Claude subscription with pi using their official Claude Code SDK.
Track complete prompts sent to the LLM gateway — system prompt, tools, messages, responses, and usage
CodeCompass cut 66%-82% of token usage on real coding work. And used to map your codebase and reduce the amount of code your agent reads and opens.
Pi web search tool that auto-detects available API keys and routes to the right provider
Claude Usage Monitor TUI - monitor your Claude usage in real-time
Pi extensions for llama.cpp router — auto model discovery and tokens/second display
Sakana Fugu provider extension for the Pi agentic harness (OpenAI-compatible, /login support)
Extra LLM models for pi (OpenAI Codex GPT-5.5).
pi (pi.dev) extension that registers the Salesforce LLM Gateway as a model provider.
Pi extension: register Volcengine Ark (火山引擎方舟) models as a custom provider.
Cursor OAuth provider extension for pi-coding-agent
Cursor agent provider extension for pi — use Cursor's models and agent directly from pi
Pi extension for logging into and switching between multiple OpenAI Codex accounts
Pi extension that auto-discovers models from an Olla gateway endpoint and registers them as a provider.
TensorX.ai API provider extension for pi coding agent.
OpenCode provider extension for pi coding agent — Zen and Go models
Provider-agnostic ultraplan engine for Pi — fan-out planning, adversarial verification, structured plans, worktree-isolated execution. Reference adapter: fairy-tales-deepseek-openai / deepseek-v4-pro.
pi extension: /usage aggregate panel with Kimi quota and local pi session consumption
A pi extension that displays active model provider subscription quota usage in the status bar. Supports z.ai (GLM) and OpenAI Codex with provider-specific quota percentages, plan levels, and reset times. Updates every 60 seconds during agent runs.
Pi extension: Kimchi (kimchi-dev) provider with browser login for /login
Thermos plugin for Pi with provider detection for pi-subagents implementations.
Pi package that adds [CoreInfra AI Hub](https://hub.coreinfra.ai/) as a model provider.
Pi extension for llama.cpp integration. Supports router, single and legacy models. Supports multiple servers.
Openference provider for pi — OpenAI-compatible endpoint with /login flow and live model discovery
Pi extension for quickly switching between multiple accounts/API keys per provider.
Pi extension scaffolding for showing subscription and quota usage across AI providers.
Pi extension that adds web_search and web_fetch tools using each provider's native search backend (ZAI MCP, Anthropic, Google, OpenAI, xAI, Claude Code subscription via claude-bridge), with DuckDuckGo as a universal fallback.
Cloud cost awareness plugin for coding agents (Claude Code, Codex, Copilot, Qoder, Gemini CLI, OpenCode, pi, Cursor, Windsurf, Cline, Kiro)
FriendliAI model provider for Pi with live model discovery
Pi extension: keep agent sessions alive across provider rate-limits, quota exhaustion, and transient auth failures by re-injecting continuation messages on an escalating retry schedule.
Agent evaluation harness. Judge sessions on success, tool usage, efficiency, methodology. Inspired by opencc.
Pi extension package that rewrites openai-responses payloads to use top-level instructions instead of system/developer input messages.
Deterministic Codex model and thinking-level router for Pi
A Pi extension that routes agents based on task descriptions using useWhen and avoidWhen criteria