Running an AI Company for EUR 0.024 in External Spend

Running a company powered entirely by artificial intelligence might sound like an expensive futuristic enterprise, but for one developer managing six autonomous AI agents, the total metered external spend has stopped at a mere EUR 0.024. Every single line item, what it purchased, and the caveats behind this figure reveal both the beauty of cheap model costs and the illusions they can create.

The entire bill consists of six calls, each costing EUR 0.004. Every call was an image generation executed by the builder agent, sequenced 35 through 40 in the spend ledger, and timestamped between 20:47:53Z and 20:48:12Z on October 12, 2026. Nineteen seconds of spending, all dedicated to producing six robot avatars for an internal team standings page.

Stock photo for illustration only, not from the actual event

Everything else the company accomplished—including 551 cards, 540 completions, 722 comments, 1,171 attached artifacts, and every public page read across marketplaces and registers—cost EUR 0, as reading data incurs no charge. The underlying mechanics of these numbers prove just as intriguing as the grand total.

While multiplying six lines at EUR 0.004 yields EUR 0.024, the single largest cost line in the ledger is not among those successful generations. Instead, it is an estimated EUR 0.03 spend on a request that returned a 401 unauthorized error. Providers can bill for failed requests, and had the budget been calculated solely based on successful calls, that EUR 0.03—surpassing the company's entire successful spend—would have gone completely unnoticed. Failure carries a price that few business models account for.

"Failure has a price, and it is usually the line item nobody models."

AI Company Creator

Out of 43 ledger decisions, 35 were denials, mostly due to malformed credentials including the creator's own setup attempts. The gate functioned correctly every time, proving that a strict security gate is an inexpensive lesson against larger errors.