Artificial Intelligence (AI) has become a core infrastructure component of modern software engineering, reshaping everything from how code is written to how software teams deliver value. As AI developer tools move from experimental add-ons to a mission-critical capability, engineering leaders face increasing pressure to distinguish genuine productivity gains from marketing hype. This report benchmarks the rapidly evolving landscape of AI coding assistants, agents, and AI-native IDEs – hereby referred to as AI developer tools. By doing this, we provide a clear path to understanding not just which tools are leading, but how they are fundamentally changing how developers work. The goal is to highlight for vendors where the strongest opportunities and gaps exist.

To understand the success of the AI developer tools included in our research, we use different metrics, with the main ones being adoption – the percentage of developers in our survey who are currently using each tool – and satisfaction (CSAT) – how developers score the help that they receive from each tool on certain tasks, equal to the proportion of four- and five-star reviews. To supplement these, we examine other metrics, like workflow reliance, task-level quality, trustworthiness, and measurable productivity gains.

The results in this report were collected in Q1 2026 from a global panel of 2,393 professional developers who use AI developer tools in their workflows. We benchmarked 20 of the most prominent AI developer tools, selected for their market impact and technological relevance. They are, in alphabetical order: Aider, Amazon Q Developer, Amp (formerly Sourcegraph Amp), Claude Code, Cline, Cursor, Devin, Gemini CLI / Gemini Code Assist, GitHub Copilot, GitLab Duo, Google Antigravity, JetBrains AI, Kiro, Mistral Code, OpenAI Codex, Replit, Tabnine, Trae, v0 by Vercel, and Windsurf.