Hi, my name is Tom Smykowski, I'm a staff full-stack engineer. I build and scale SaaS platforms to millions of users, working end-to-end from system architecture to frontend to mobile. On this blog I share what I learn about AI-assisted development and building products that actually work for people.
What This Article Covers
Explore a detailed benchmark comparing several AI models, including Claude, GPT-4.1, Gemini, and SWE-1, as they attempt to build a simple Vue TODO application using the Windsurf platform. This article delves into how each AI model handles tasks such as setting up a project, managing dependencies, and adhering to modern coding practices, highlighting both their strengths and weaknesses.
Questions This Article Answers
- Which AI model is the fastest at building a Vue TODO app in Windsurf?
- How do different AI models manage common development tasks and challenges?
- What coding practices do the AI models struggle with, and how does this impact the final application?
- Are there any noticeable patterns in the AI models' handling of CSS and API usage?
- What insights can be drawn from the performance of these models in creating a basic app from scratch?
Length and Time
A detailed analysis with comparisons and practical insights. Approximately 12 minutes to read.