October 28, 2025
Zurzai.com is a health tracking app I built on V0, hosted on Vercel, with a Supabase backend. It helps you track fitness, calories, weight, notes, and mindfulness—then uses AI to analyze your data and give you personalized feedback.
The Performance Problem
Initially, I hooked zurzai.com up to the OpenAI API. It was easy to integrate, but I quickly noticed something frustrating: the OpenAI API was very slow, especially for complex queries analyzing multiple days of health data.
The Experiment
I decided to run an experiment and switch to the Claude API to see if performance would improve. The migration wasn't perfect on the first try—but it worked correctly on the second attempt. And the results? Claude is noticeably faster, on the order of 3 to 4 times as fast, especially for those complex queries that analyze multiple dimensions of health data.
Easy to Switch Providers
One of the benefits of this tech stack (V0 + Vercel + Supabase) is that it's very easy to switch AI providers. You're not locked into one vendor's API. This flexibility means you can optimize for performance, cost, or capabilities depending on your needs.
Same Capabilities, Better Performance
The Claude API handles all the same analysis that the OpenAI API was doing—tracking fitness trends, analyzing calorie intake, monitoring weight changes, reviewing mindfulness notes, and providing personalized recommendations. The difference is that it does it significantly faster.
The Bottom Line
For zurzai.com, switching to Claude API was a clear win. If you're building AI-powered apps and performance matters—especially for complex queries—it's worth testing Claude API alongside OpenAI to see which performs better for your use case.
Try Zurzai.com yourself and experience the faster Claude-powered analysis firsthand.
← Back to Blog