Pricing

AutoFunnel started as an internal tool, running our own marketing on our own AI provider accounts. That detail matters more than it sounds like it should: every generation, every test, every rewritten angle came out of a budget we watched ourselves.

That constraint shaped the product from day one. It had to test freely — try ten angles to find the one that converts, throw away nine — without someone eyeballing a usage meter and flinching. Restraint kills the exact behavior that makes the loop work.

Once it was running two brands, then helping a few people outside the company, it became obvious this wasn't just a convenience — it was doing a job most businesses pay an agency, a freelancer, or their own exhausted hours for. That's when "sell it" stopped being a strange idea.

The obvious move was tiers: a starter plan with limited generations, a pro plan with more, an enterprise plan with "unlimited." We tried sketching it and kept hitting the same wall — every version capped exactly the behavior that made the tool worth using in the first place.

A package forces a ceiling on testing. Fewer angles tried, fewer variants run, fewer cycles of the loop per month — precisely the opposite of what compounds results. Selling a capped version of a tool built on uncapped iteration would have meant selling something worse than what we actually use ourselves.

Bring your own key: you connect your own accounts with OpenAI, Anthropic, Grok, or whichever provider you prefer, and every generation runs on your key, at your provider's actual rate. We don't meter usage, and we don't stand between you and the bill.

No resale margin. You pay the provider's real rate — the same rate we pay ourselves — instead of a marked-up allotment of "credits."

No artificial ceiling. Test twenty angles in a day if the week calls for it. The only limit is the one you set.

Full visibility. Your provider bill is the ground truth. Nothing is obscured behind a subscription tier's fine print.

Freedom to switch. Prefer a different model for a specific job — better at long-form, cheaper for volume, faster for iteration? Swap it. Nothing locks you into one provider's roadmap.