Aider Open source

What it is

Aider is a long-running open-source (Apache-2.0) project for pair programming with a large language model in your terminal. You start it in a git repository, add the files you want to work on, and describe the change; Aider edits the files and commits each change to git with a descriptive message, so every step can be reviewed with normal git tools or undone.

The README's install method is python -m pip install aider-install followed by aider-install. It works with cloud models from Anthropic, OpenAI, DeepSeek and others, and with local models, using your own API keys.

Key features

• A repository map that gives the model a compact view of the whole codebase, which helps on larger projects
• Automatic git commits for every change
• Support for 100+ programming languages
• Runs linters and tests after edits and can fix the problems they report
• Watch mode: leave instructions as comments in your own editor and Aider picks them up
• Voice input, and images or web pages added as context
• A copy/paste mode for working with LLM web chat interfaces

Skills, subagents and MCP

Aider does not document support for MCP servers, Agent Skills or subagents. Project conventions are handled with plain files instead: you can load a conventions file read-only with aider --read CONVENTIONS.md or the in-chat /read command, or list it under read: in .aider.conf.yml so it loads every time. Loading it read-only also lets it be cached when prompt caching is available.

Pricing

Aider is free and open source. Your only cost is the model provider you connect, billed at their API rates, or nothing if you run a local model.

Who it is for

Developers who want tight control over each change, value clean git history, or want to use a specific or local model. It also suits people who prefer to stay in their own editor and use the terminal alongside it.

Verdict

Aider is mature, predictable and well documented, and its git-first approach makes changes easy to audit. It is more hands-on than autonomous agents and lacks MCP, skills and subagents, so it is the wrong pick if you want to plug into that ecosystem. Development activity has also slowed compared with the fastest-moving agents in this directory.