# Aider — AI pair programming in your terminal, an open-source CLI agent that edits your code and commits to git.

> Source: The Agents Index — https://theagentsindex.com/aider (structured, researched, re-verified)
> Facts last verified: 2026-07-23

Aider is an open-source, terminal-based AI pair programmer, released under the Apache-2.0 licence. You run it in your shell inside a git repository and describe changes in natural language; Aider edits the relevant files and — distinctively — commits each change to git with a sensible message, so every AI edit is a reviewable, revertible commit. It is bring-your-own-key and model-agnostic, working with Anthropic, OpenAI, DeepSeek, local and effectively any LLM. Free and widely adopted (47,000+ GitHub stars and 6.8 million installs), it is a favourite of developers who live in the terminal and want AI edits kept under version control.

| Fact | Value |
| --- | --- |
| Website | https://aider.chat |
| Best for | Terminal-native developers who want free, model-agnostic AI edits that land as clean git commits. |
| Not for | Developers who want a graphical in-editor experience or a fully autonomous, hands-off cloud agent. |

## Verdict

Aider is the definitive terminal AI pair programmer: open-source under Apache-2.0, model-agnostic, and built around a git-native workflow where every AI edit becomes a reviewable, revertible commit. That commit discipline plus its repository map and auto lint/test fixing make it one of the safest ways to let a model touch a real codebase, and being bring-your-own-key keeps it free and flexible — route work to Claude, GPT, DeepSeek or a local model as you like. Its well-known code-editing leaderboards are a genuinely useful, oft-cited signal on model ability. The trade-offs are that it is terminal-only with a steeper on-ramp than a GUI tool, you pay your provider per token, and it is an interactive pair programmer rather than a fire-and-forget agent. Pick Aider if you live in the terminal and want free, auditable, model-agnostic AI edits under git; skip it if you want a graphical editor or a fully autonomous cloud agent.

## Strengths and weaknesses

- ✓ Completely free and open-source (Apache-2.0) — pay only for the model you bring.
- ✓ Git-native commit workflow makes AI edits auditable and easy to undo.
- ✓ Model-agnostic with a respected public leaderboard on model coding ability.
- ✗ Primarily terminal-only and text-driven — there is an experimental `--browser` web UI, but no polished graphical editor, and a steeper learning curve than an in-editor tool for non-CLI users.
- ✗ You pay per token to your model provider, and results depend entirely on the model you supply.
- ✗ Less “autonomous” than agents like Devin — it is an interactive pair programmer you drive, not a fire-and-forget agent.
- ⚠ It is an interactive CLI pair programmer, not a headless autonomous agent or a hosted service.
- ⚠ No bundled model or managed support; you own key management and model choice.

## Key features

- **Git-native commits** — Every AI edit lands as a git commit with a sensible message, so changes are reviewable and revertible by default.
- **Repository map** — Sends a compact Tree-sitter map of the whole codebase so the model has context on large projects without dumping every file.
- **Architect/editor mode** — Splits a hard task across two models — one proposes the solution, a second turns it into precise file edits — for better results at the cost of an extra API call.
- **Model-agnostic, BYO key** — Works with Claude, GPT, DeepSeek, local and nearly any LLM; you supply the key and pay the provider directly.
- **Auto lint & test fixes** — Can run your linters and tests after an edit and automatically fix what it broke.
- **Public model leaderboards** — Maintains widely-cited code-editing and polyglot benchmarks of how well models perform its edit format.

## Use cases

- **Terminal-first development** — A developer who lives in the shell makes AI edits without leaving the terminal or their existing editor.
- **Version-controlled AI edits** — A team keeps every AI change as a clean, revertible git commit for easy review.
- **Model evaluation** — An engineer uses Aider’s edit format and leaderboard to compare how models actually perform on code edits.

## Integrations

Git (native commits) · Model providers — Anthropic, OpenAI, DeepSeek, local/Ollama · IDEs via watch mode (edit-from-comments) · Linters & test runners

## Sources

- https://aider.chat
- https://github.com/Aider-AI/aider
- https://aider.chat/docs/usage/modes.html
- https://aider.chat/docs/usage/browser.html
- https://codegen.com/ai-tools/aider/