lipcode

Fast, cost-optimized AI coding CLI built in Rust. Local-first with automatic cloud fallback.

$ curl -fsSL https://raw.githubusercontent.com/jongha/lipcode/main/install.sh | bash click to copy
Copied!

Features

⚡

Fast Startup

Single Rust binary. Near-instant launch with no runtime dependencies.

🏠

Local First

Runs on Ollama out of the box. Your code never leaves your machine.

💰

Cost-Aware Routing

Automatically picks the cheapest model that fits the task complexity.

🔧

Built-in Tools

File read/write/edit, bash, grep, glob. The agent drives your codebase.

👥

Multi-Agent

Role-based agent delegation. Complex tasks are broken down and routed.

🔒

Permission System

Safety guards for destructive operations. You approve before it runs.

How It Looks

Welcome to lipcode. [qwen3-coder] Light · local > read src/main.rs and add error handling ● file_read — src/main.rs (24 lines) ● file_edit — added Result return type, wrapped IO calls Done. Added error propagation with thiserror to main.rs. RUNNING lipcode $0.00 · 1.2k tokens

Quick Setup

Local (Ollama)

Zero config. lipcode auto-detects Ollama and pulls the default model on first run.

$ ollama serve
$ lipcode

Cloud (OpenRouter)

Set your API key for access to Claude, GPT, Gemini, and more.

$ export OPENROUTER_API_KEY="sk-..."
$ lipcode

Update

Self-update to the latest release with a single command.

$ lipcode --update