Software product · Lessons learned

ZenPins

ZenPins, day 29 of a solo AI venture: $45.20 spent in 30 days, $0 revenue

On day 29 of the author's solo AI venture, the author published a 30-day ledger for the hyper-casual game ZenPins: $45.20 spent (Google Play registration $25, Claude 3.5 Sonnet API $14.20, Cloudflare domain $6) and $0.00 revenue. The product has not been publicly released.

The story

On day 29 of the author's “solo AI venture”, the author published a 30-day financial ledger for ZenPins, a hyper-casual mobile game. The amount column in the long post's table is labelled USD.

Total spend was $45.20: a $25.00 Google Play Console developer registration fee, $14.20 for the Claude 3.5 Sonnet API (used for level generation and multilingual packaging) and $6.00 for an annual Cloudflare domain. Unity and Cursor were used on free tiers and recorded as $0.00, and no Google Ads or Meta ads were run before the closed test. The author's “about 325 yuan” is their conversion of the same dollar spend, not a separate yuan expense.

Revenue was $0.00. The product has not been publicly released and has no in-app purchases yet; the author says the feel of the first 20 levels is still being tuned, and the 20-person, 14-day closed test list on Google Play is still being recruited. The author refuses to discuss LTV, CAC or ROI before the first dollar comes in.

AI was used to write multilingual packaging scripts and automatically check deadlock logic across 100 levels. In the thread someone asked whether the $14.2 API spend went mainly to testing or production runs; no answer from the author was seen, although the long post says the cost covered level generation and multilingual packaging.

Source excerpt

【一人 AI 出海实战 · Day 29】30 天真实财务账单:实际花出去的钱,与挂在账上的 0 收入 昨天在…

@re_refoto · Read the full source

What was delivered

A hyper-casual mobile game using AI to write multilingual packaging scripts and check deadlock logic across 100 levels.

Costs, risks and open questions

Not yet publicly released and no in-app purchases; level feel still being tuned; closed test list still being recruited. Author's lesson: very low cost is a moat that buys time to stay in the game, and no revenue means no LTV/CAC/ROI talk.

Tools and topics

  • Claude
  • Cursor

When citing this case, credit AI Income Journal and include a link to this page. Retain the original author's credit and source link. Content use and attribution

Report an issue with this case