Open Source CLI Tool

Andi AIRun

Executable markdown with shebang, Unix pipes, and output redirection. Extends Claude Code with cross-cloud provider switching and any-model support — local or 100+ cloud models.

~/ — bash

# Make any markdown file executable

$ cat task.md

#!/usr/bin/env ai

Analyze my codebase and summarize the architecture.

$ chmod +x task.md && ./task.md

# Pipe data in, redirect output out

$ cat data.json | ./analyze.md > results.txt

# Use Claude Code with any model

$ ai --vercel --model openai/gpt-5.2-codex

$ ai --ollama --model qwen3-coder

git clone https://github.com/andisearch/airun.git && cd airun && ./setup.sh

View on GitHub

Features

Everything you need to automate AI

Executable markdown scripting, Unix pipes, and multi-provider support — use Claude Code with any model, local or cloud.

Executable Markdown

Add a shebang line to any markdown file and run it directly. Your prompts become reusable, version-controlled tools.

#!/usr/bin/env ai

Any Model, Any Provider

Use Claude Code with any model — Claude, GPT, Gemini, Llama, Qwen, and more. Switch between 8 providers with a flag. Hit a rate limit? --resume on another.

ai --vercel --model openai/gpt-5.2-codex

Unix Pipes

Pipe data in, redirect output, chain scripts together. Full Unix semantics for automation — stdout is data, stderr is diagnostics.

cat data.json | ./analyze.md > out.txt

Local Models

Run Claude Code with any local model via Ollama or LM Studio. Free, private, no API keys. Use Qwen, Llama, DeepSeek, or any GGUF/MLX model.

ai --ollama --model qwen3-coder

Providers

One CLI, any model, every cloud

Use Claude Code with any model — locally or via 100+ cloud models on Vercel AI Gateway. Switch providers mid-conversation to bypass rate limits or run for free.

🦙

Ollama

LocalFree

--ollama

🔬

LM Studio

LocalFree

--lmstudio

☁️

AWS Bedrock

--aws

🔷

Google Vertex AI

--vertex

🔑

Anthropic API

--apikey

🟦

Microsoft Azure

--azure

▲

Vercel AI Gateway

100+ models

--vercel

✦

Claude Pro

Subscription

--pro

Usage

Simple, powerful commands

One unified CLI for everything. Model tiers, provider switching, session resume, and persistent defaults.

examples

# Run a script with any provider + model tier

$ ai --aws --opus task.md

# Run locally for free

$ ai --ollama

# Use any model via Vercel AI Gateway

$ ai --vercel --model openai/gpt-5.2-codex

# Hit a rate limit? Resume on another provider

$ ai --aws --resume

# Set your preferred default

$ ai --vercel --opus --set-default

# Chain scripts together

$ ./generate.md | ./review.md > final.txt

Quick Start

Install in 30 seconds

Requires Claude Code (Anthropic's AI coding CLI). Works on macOS, Linux, and Windows via WSL.

install

# Install Claude Code (if not already installed)

$ curl -fsSL https://claude.ai/install.sh | bash

# Install Andi AIRun

$ git clone https://github.com/andisearch/airun.git

$ cd airun && ./setup.sh

# Start using it

$ ai task.md