Fluid Compute is the default Claude-hosting runtime on Vercel

What changed: Vercel's Fluid Compute reuses function instances across concurrent requests, drastically cutting cold starts. It's now the default for new Claude-powered apps built on Vercel (including this training packet's AI tutor).

Why it matters for us: the tutor sidebar gets faster under load — fewer spinners, more streaming start-of-response. No code change needed on our side; the Vercel runtime handles it.

Prompt to try:

On your next lesson, open the tutor and ask three questions in a row. Notice the second and third responses start streaming faster than the first.