# Aider AI pair programming in your terminal that maps your codebase, auto-commits to git, and works with 100+ languages and any LLM provider. ## Summary Aider is an open-source terminal tool for AI pair programming that works with any LLM provider, maps your codebase with a repository map, and automatically commits changes to git. It supports over 100 programming languages and integrates with local and cloud models through a flexible provider system. ## Guide Aider is one of the original terminal coding agents, and it remains one of the most popular. Its model-agnostic design and automatic git integration make it a practical choice for developers who want AI pair programming without vendor lock-in. ### What it is Aider is an open-source terminal tool for AI pair programming. It connects to any LLM provider, builds a map of your codebase, and works with you to edit code through natural conversation. Every change is automatically committed to git with a descriptive message, giving you a clean audit trail. ### Why it matters Most coding agents lock you into a specific model provider. Aider's model-agnostic design means you can use the best model for each task — GPT-4 for complex reasoning, Claude for writing, or a local model for privacy-sensitive code. Combined with automatic git integration, it's one of the most production-ready coding agents available. ### How it works Install via pip and run 'aider' in any git repository. Aider scans your codebase and builds a repository map that helps the model understand file relationships. You then chat naturally about what you want to change. Aider edits files, shows diffs, and auto-commits when you're satisfied. You can undo any commit, review the AI's reasoning, and iterate until the code is right. ## Use Cases - Adding tests to existing code: Point Aider at a module and ask it to generate unit tests. It reads the code, understands the API, and creates tests that match the existing patterns. - Refactoring with confidence: Describe the refactoring you want, review Aider's diffs, and commit only when you're satisfied. Every step is tracked in git. - Cross-language development: Aider supports 100+ languages, so you can use the same workflow for Python backend code, TypeScript frontend, and infrastructure scripts. ## Alternatives - Use Claude Code for deeper codebase reasoning vs Claude Code: Aider offers model flexibility and git integration. Claude Code may handle complex multi-file reasoning better but is limited to Anthropic models. - Use Codex CLI for faster, lighter-weight tasks vs Codex CLI: Codex CLI is faster for quick edits. Aider is more feature-complete with better git integration and multi-model support. ### Getting Started - Read the documentation: https://aider.chat/docs - Inspect the repository: https://github.com/Aider-AI/aider ### FAQ - Is Aider open source? - Yes. Aider is released under the Apache-2.0 license, making it fully open source for inspection, modification, and redistribution. - Who should use Aider? - Developers who want AI pair programming with the flexibility to use any LLM provider, especially those who value automatic git integration and clean commit history. - How does Aider compare to other coding agents? - Aider's main differentiators are its model-agnostic design and deep git integration. It's more interactive than fully autonomous agents but offers more control and flexibility. ## Why It Matters Aider is one of the most established terminal coding agents, with a large community and a proven track record of handling real-world codebases. Its model-agnostic design means developers aren't locked into any single AI provider, and its git integration makes it one of the few agents that treats version control as a first-class concern. ## Best For - Developers who want model-agnostic AI pair programming in the terminal - Engineers who value automatic git integration and clean commit history - Teams experimenting with different LLM providers for coding tasks ## Not For - Users who prefer GUI-based coding environments - Developers who want a fully autonomous agent that plans and executes multi-step workflows without interaction ## What It Actually Does - Model-agnostic architecture: Aider works with any LLM provider — OpenAI, Anthropic, Google, local models, and more — through a flexible configuration system. - Why it matters: Developers can switch models based on task, cost, or speed requirements without changing their workflow. - Automatic git integration: Aider automatically commits changes with meaningful commit messages, creating a clean audit trail of AI-assisted edits. - Why it matters: Git integration means every AI edit is tracked, reviewable, and reversible — critical for production use. - Repository map for context: Aider builds a map of your codebase structure, helping the model understand relationships between files and modules. - Why it matters: Better context leads to more accurate edits, especially in large codebases where understanding architecture matters. ## Typical Use Cases - Iterative feature development: Work with Aider in a conversational loop: describe the feature, review edits, request changes, and commit when satisfied. - Multi-file refactoring: Use Aider's repository map to refactor across multiple files while maintaining consistency and auto-committing changes. - Learning and exploration: Ask Aider to explain code, suggest improvements, or walk through unfamiliar modules while it tracks all changes in git. ## How It Compares - Choose Aider for model flexibility and git integration vs Claude Code: Aider supports 100+ LLM providers and has deeper git integration. Claude Code may handle complex reasoning better but locks you into Anthropic's models. ## Command Line ### Install Aider Install via pip, then run 'aider' in any git repository. Configure your preferred LLM provider in the settings. ```bash pip install aider-chat ``` ## Facts - Category: agents - Resource type: agent - Open source: yes - License: Apache-2.0 - Last verified: 2026-05-27 - GitHub repo: Aider-AI/aider - GitHub stars: 45400 ## Capabilities - workflow-orchestration ## Structured Use Case Tags - developer-workflow ## Getting Started - Open the GitHub repository: https://github.com/Aider-AI/aider - Read the documentation: https://aider.chat/docs - Visit the project website: https://aider.chat ## Links - GitHub: https://github.com/Aider-AI/aider - Homepage: https://aider.chat - Docs: https://aider.chat/docs ## Structured Outputs - JSON: https://www.openagent.bot/agents/aider.json - Markdown: https://www.openagent.bot/agents/aider.md - Canonical: https://www.openagent.bot/agents/aider