Aider
Terminal AI pair programmer that edits files in your git repo using the LLM of your choice, committing each change.
Aider edits your local codebase with the LLM you choose and commits each change to git.
Aider is a command-line tool that lets you pair program with a large language model against your own repository. It suits developers comfortable in a terminal who want to pick their own model and keep every AI edit visible in git. You bring the API key or a local model.
What it is
You run aider inside a project, add files to the chat, and describe changes in plain language. It builds a map of your whole codebase to work in larger projects, applies edits to your files, and commits them with generated messages. It can also lint and run your tests after each change and try to fix what fails.
Who it's for
It suits developers who live in the terminal and want model choice, including local models, plus git history as the safety net. It does not suit people who want a graphical editor with built-in model access, because there is no bundled model and costs come from your API provider.
Key features
Codebase map
Builds a map of your entire repository so the model works better in larger projects.
Automatic git commits
Commits each change with a sensible message so you can diff and undo AI edits with normal git tools.
Lint and test loop
Runs your linters and test suites after changes and can fix problems they report.
Almost any LLM
Connects to cloud models or local ones, with named shortcuts for Claude, DeepSeek and OpenAI models.
Watch mode for IDEs
Ask for changes by adding comments in your editor and aider picks them up.
Images, web pages, voice
Add screenshots or URLs to the chat, or speak your requests.
Up and running in 3 steps
Our honest verdict
What's great
- Every edit becomes a git commit, so reviewing and reverting AI changes uses tools you already know.
- You choose the model per run, including local models, rather than being tied to one vendor.
- The copy/paste-to-web-chat mode lets you work with an LLM through its browser chat instead of an API.
- It runs from your own editor workflow through comment-triggered watch mode, with no separate IDE to adopt.
Where it falls short
- You pay your model provider for every request, and the README recommends commercial models; costs depend on how much code you send.
- It is terminal-only, so you need to be comfortable with Python packaging, git and command-line flags.
- The README's recommended-model list names Claude 3.7 Sonnet, o3-mini and GPT-4o, so check the docs for which current models it supports well.
- Result quality depends on the model you connect, so you still need to review each diff it commits.
Pricing & license
Aider is free under the Apache-2.0 licence and the project's site shows no paid plan. The cost is your LLM provider's API charges, or the hardware to run a local model.
Open remote AI-training roles
860 open remote AI-training roles, updated November 14, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
These reviewed AI tools do the same job instantly — from our scored directory.
Use it, or have us set it up
FAQ
Is Aider free?
The tool itself is free under Apache-2.0, but it calls an LLM you must pay for unless you run a local model. The site shows no paid Aider plan.
Can I use local models?
Yes. The README says it can connect to almost any LLM including local models, and the docs list Ollama, LM Studio and OpenAI-compatible APIs.
Does it work with my editor?
It runs in the terminal beside your editor, and its watch mode reads AI comments you add in your IDE.
