The only unpaid tier

Don't kill your free tier. Demote it. Fold the free plan into an earned tier as a small baseline that keeps working for everyone, and make posting the way past it.

Killing the free plan stops the compute bill and the top of your funnel in the same move. Instead, make it work for you.

Why founders reach for the kill switch

The case writes itself, especially for AI products. Every free request costs real inference. Anonymous signups invite multi-account abuse. And the funnel is silent: free users arrive, use the product, and tell no one.

What killing it actually costs

Deletion has a bill too, and it arrives later. "Try it free" is the strongest call to action a self-serve product has, and every alternative converts colder traffic worse.

A time-boxed trial swaps a known population of free users for dropouts you never hear from again. And a product with no unpaid path stops being recommendable: "just try it" needs somewhere to send people.

The demotion

Keep a small baseline, enough that the product does its job for everyone. Everything above it is earned or paid. Earning means a public, disclosed, verified post about the product; the product sets what a post is worth and how often one user can earn.

The framing matters. This is not a tier stacked on top of free, which reads as a growth hack. It is the unpaid tier, singular. Nothing is taken away; a way up is added.

How small should the baseline be?

Small enough that someone with real usage feels the wall in their first week or two. That is when a new user is most engaged and has the freshest reason to post. A baseline nobody ever hits is the old free tier with a new name.

This is not hypothetical for us: Earnesty runs its own earned tier, and the running programs and precedents are their own page.

Questions founders ask