Aider is an open-source terminal-based AI pair programming tool created by Paul Gauthier and the Aider-AI community. It operates directly inside local Git repositories, interacting with your code files to implement features, fix bugs, and refactor existing modules. In daily development workflows, Aider runs as an interactive terminal CLI session. Developers supply prompts and reference files, and Aider automatically applies edits directly to source files on disk. It supports major commercial and open-source models, including Anthropic Claude 3.7 Sonnet, OpenAI o1/o3-mini/GPT-4o, DeepSeek R1/Chat V3, and locally hosted models via standard OpenAI-compatible endpoints. Aider distinguishes itself through its repository map architecture, which uses tree-sitter parsing to create an index of definitions and signatures across the entire project. This provides LLMs with cross-file awareness while conserving token context. In addition, Aider strictly integrates with Git, automatically writing concise commit messages for each change and providing feedback loops via automated test and lint execution. The tool is distributed under an open-source license via PyPI and GitHub. Users bring their own API keys, paying standard token pricing directly to model providers with no intermediary platform fee.
Repository map — Indexes your full codebase using tree-sitter to give models broad context on large projects while optimizing token usage. - Automated Git integration — Modifies files on disk and automatically commits changes with descriptive commit messages. - Multi-model & local LLM support — Connects to Claude 3.7 Sonnet, DeepSeek R1/V3, OpenAI o3-mini/GPT-4o, and local models via standard API endpoints. - Automated lint and test loops — Runs configured test suites and linters after code modifications, automatically prompting the LLM to fix any failures. - IDE watch mode — Monitors codebase files for instructions written as inline comments, allowing developers to drive Aider directly from any text editor. - Multimodal & URL context — Ingests images (such as UI mockups or bug screenshots) and web documentation URLs directly into the chat session. - Voice-to-code — Accepts spoken audio prompts in the terminal to request refactors, tests, or bug fixes hands-free.
Use cases
Use Case: Test-Driven Bug Fixing A developer runs Aider with their test suite command configured, prompting the agent to reproduce an issue and fix it. Aider updates the code, runs the test suite, reads the failure output, and iterates on the implementation until all tests pass before making a Git commit.
Use Case: Cross-File Feature Implementation When implementing a feature that spans database models, routing layers, and template files, Aider uses its codebase map to locate cross-file references, applies all necessary edits across the target files, and packages the work into an atomic commit.
Use Case: Editor-Agnostic Comment Automation A developer working inside Vim, Emacs, or VS Code runs Aider in watch mode and writes prompt comments directly into the source code. Aider detects the file changes, executes the requested code generation, and replaces the instruction comment with the implementation.
Who it is for
Aider is designed for professional software engineers who prefer terminal workflows, rely on Git for version control, and want direct control over model selection and API keys. It is not intended for non-developers seeking GUI-based no-code app generators or teams seeking managed enterprise SaaS interfaces.
Install guide
Installation Install Aider using the official installer package or directly via pip: bash python -m pip install aider-install aider-install
First Run Navigate to your project repository and launch Aider with your preferred model and API key: bash cd /path/to/your/project # Using Claude 3.7 Sonnet export ANTHROPIC_API_KEY="sk-ant-..." aider --model sonnet # Or using DeepSeek / OpenAI directly aider --model deepseek --api-key deepseek=your_key aider --model o3-mini --api-key openai=your_key
Example Session Launch Aider with relevant files and issue an interactive prompt: text $ aider app/routes.py tests/test_routes.py app/routes.py tests/test_routes.py > Add a health check endpoint at /healthz returning JSON {status: 'ok'} and write a corresponding unit test. Aider will generate the changes, modify both files, run any configured test commands, and create a Git commit automatically.
GitHub Copilot CLI — GitHub Copilot CLI is a terminal-based AI coding assistant developed by GitHub. It integrates the Copilot coding agent runtime directly…
Agentty — Agentty is an open-source, high-performance native terminal-based AI coding agent written in C++ by 1ay1. Engineered specifically for speed and…
Terminai — Terminai (Termin.AI) is an open-source terminal process manager and AI agent harness developed by emosenkis. Built on top of the…
Amp — Amp (developed by AmpCode) is an autonomous coding agent platform and remote development harness designed to execute software engineering tasks…
Cline — Cline is an open-source autonomous coding agent runtime maintained by Cline Bot Inc. and an active open-source community. Available as…
Continue — Continue is an open-source AI code assistant and agent framework originally developed by the Continue team (continuedev) and recently acquired…