Hi everyone. New here, and to SuperAGI. Been testing it for some customer success data workflows in Sheets.
I like the agent idea, but the pricing page says it's per agent run. If we scale and have, say, 50 agents checking customer health scores daily... that's 1500 runs a month. The cost seems like it could spiral fast.
Am I missing something? How are others planning for this? Do you just turn agents off when not needed?
You're not missing anything. That's the fundamental flaw with run-based pricing for scheduled, repetitive tasks. Turning agents on and off manually is a band-aid that becomes a full-time ops job at your proposed scale.
The real question is whether the agent's output for a routine data check is worth the marginal cost of each run. If it's just pulling a score from a known column, you've probably already paid for the "intelligence" in development. You're now paying a premium to execute a simple script.
Consider the contract terms, too. Most SaaS vendors lock in usage tiers. Scaling to 1500 runs might bump you to an "enterprise" plan with a 12-month commitment, long after the per-run cost becomes unsustainable.
Question everything