Shutting down my SaaS because it depended too much on Claude/Codex

I’m shutting down buildspeedy.

It was a tool I built to help people build mobile apps faster and cheaper by using Claude and Codex subscriptions as part of the workflow.

The main problem: Build Speedy relied heavily on claude -p to make the app-building process fast and affordable. With Claude moving this usage toward extra credits instead of being covered by the subscription, the economics of the product changed a lot.

Technically, I could probably work around it with other solutions. But that made me realize a bigger issue: the product was too dependent on Claude/Codex, and the app itself was getting too complex to maintain for the value it provided.

So I decided to shut it down and shift my focus to B2C apps instead.

Main lessons for me:

1. If your SaaS depends too much on another platform’s pricing, your business model can change overnight.
2. “Possible workaround” doesn’t always mean “worth continuing.”
3. Complexity compounds fast, especially when the core value depends on tools you don’t control.
4. Sometimes shutting down is the right move so you can focus on a better opportunity.

Curious if anyone here has gone through something similar: shutting down a SaaS because the platform/API/pricing it depended on changed?

Author: According_Scar3032