AI for Developers

GitHub Copilot

Surely the most widely adopted AI coding assistant and deeply integrated into VS Code and Github. With cool agentic features as well. More or less the default entry point for AI-assistend coding.

Cursor

Strong AI-native code editor that works smoothely especially for larger complex software projects. Lots of agentic features as well. High usage among professional devs.

Aider

Almost seven million installs (maybe more at the time you read this) speak for themselves. Open source, terminal-based pair programming with full control and without a SaaS subscription.

Cline

„One agent across your IDE, terminal, and SDK“ is the promise, and Cline delivers just fine on it. Open source, integration with almost all major model providers and full control over everything.

Claude Code

Terminal-native agentic coding tool. Large context window. Highly agentic.

Devin, the AI software engineer

Cognition’s agentic IDE. Formerly known as Windsurf. Powerful for complex multi-repo projects and the Agent Command Center is great.

Amp

Sourcegraph’s standalone coding agent, spun off as its own product. Free to use when bringing own compute and keys.

Amazon Q Developer

AWS-native AI coding assistant. If you look for a tight integration into Lambda, CDK and so on and build on the AWS stack, this works great.

Ollama

Open-source, and you can run open-weight LLMs locally on your own machine. Very transparent pricing, very flexible and fast setup.

Philipp’s pick: OpenCode

Open-source (MIT). Terminal-native AI coding agent. Connects to almost every model, including local models. Terminal interface, desktop app, IDE extension. LSP-ready. And so on. Highly effective in all scenarios.