This entry is compiled from primary sources: Aider's official product documentation, release notes, and published licensing repositories.
It has not been hands-on tested by us. Where a limit, a price or a requirement is stated below, it is the published figure at the time of writing rather than a measured result.
Anything we have not verified is left out instead of guessed at.
Aider Review (2026) — Open Source CLI Pair Programming & API Model Costs
Apache 2.0 Open SourceCLI ToolPaul Gauthier
Aider is an open-source command-line AI pair programmer that pairs directly with Git in your terminal. Created by Paul Gauthier, Aider allows developers to edit multi-file codebases, fix bugs, write unit tests, and refactor code using natural language prompts.
Because Aider is a local CLI tool, the software itself is 100% free and open-source under the Apache 2.0 licence. Your only operational cost depends on whether you connect cloud LLM API keys or local open-weights models.
What you get with Aider (100% Free Software)
Aider can be installed in seconds via pip install aider-chat or pipx on macOS, Linux, and Windows. Core features include:
- Automatic Git Commits: Every time Aider edits code across your files, it automatically creates a clean Git commit with a descriptive commit message.
- Multi-File Codebase Context: Map repository files automatically using tree-sitter AST parsing so the model understands function references across your codebase.
- 100% Free Offline Operation (Local Models): Connect Aider to Ollama or LM Studio to run models like DeepSeek, Qwen 2.5 Coder, or Llama 3 locally for zero cost.
- Commercial API Flexibility: Supply your own API keys for Anthropic (Claude 3.5 Sonnet), OpenAI (GPT-4o), or DeepSeek R1/V3 to pay strictly for raw token usage.
- Automated Test & Lint Fixing: Pass your test runner command (e.g.
aider --test pytest) and Aider will automatically run tests and fix failing assertions.
Understanding Aider's model & API cost structure
Aider charges no subscription fees. Cost considerations depend entirely on your inference model choice:
- Local LLMs (Ollama / LM Studio): 100% free. Requires a decent GPU (e.g. 8GB+ VRAM) to achieve responsive multi-file editing speeds.
- Cloud LLM API Keys (Anthropic / OpenAI / DeepSeek): Pay per token consumed. Using DeepSeek R1/V3 keeps costs to pennies per session; heavy Claude 3.5 Sonnet usage typically costs $1–$5 per day of heavy refactoring.
Pros
- 100% free and open-source software (Apache 2.0) with zero telemetry or account requirements.
- Brilliant automatic Git commit workflow that ensures every AI change is isolated and revertible.
- Supports 100% offline private coding with local Ollama model backends.
- Automated test execution loop that runs your test suite and repairs failing code automatically.
Cons
- Terminal CLI interface — lacks a graphical side-panel UI for developers who prefer mouse navigation.
- Using top commercial models (Claude 3.5 Sonnet) requires managing API keys and usage billing.
- Local model performance relies on local workstation hardware capacity.
Best alternatives
- Continue.dev — Open-source IDE extension bringing AI chat and local model support into VS Code and JetBrains (code).
- Cursor — AI-first graphical editor fork of VS Code with visual multi-file editing (code).
- Codeium — Free hosted IDE extension providing unlimited autocomplete (code).
FAQ
Is Aider free and open source? Yes. Aider is 100% free and open-source software released under the Apache 2.0 licence.
Does Aider require paid API keys to use? No. While Aider works exceptionally well with Claude 3.5 Sonnet or GPT-4o API keys, you can run it 100% free using local models via Ollama or LM Studio.
How does Aider integrate with Git? Aider automatically tracks your project's Git repository structure and creates individual git commits with descriptive messages after applying code changes.