Pay stub: what an AI-generated track earns once it reaches Spotify

The pitch is simple: type a prompt, get a finished track, upload it, collect royalties. The stub is less simple. One AI-generated song has two fixed costs, a royalty floor it must clear before it earns a cent, and a per-stream rate small enough that the floor decides most outcomes. Here is the whole calculation, with every assumption stated so you can change it.

Assumptions

1. Generator: Suno Pro at $8 a month. Suno's pricing page lists Pro at $8 a month (less on yearly billing), with 2,500 credits, 20 song downloads a month and commercial use rights; the free plan has no commercial rights (Suno). We assume one month's subscription for one track. For how Suno compares with other generators, see our Suno vs Udio vs ElevenMusic comparison.
2. Distributor: $24.99 a year, 100% of royalties kept. DistroKid's cheapest plan, Musician, is $24.99 billed annually and says "Keep 100% of earnings" (DistroKid). TuneCore's cheapest plans are also $24.99 a year, with a 20% fee named only for social-platform income (TuneCore). More on choosing one in our distributor guide.
3. Per-stream rate: $0.003 to $0.005. A third-party estimate attributed to the Duetti Music Economics Report and Ditto data, before distributor fees (Unchained Music). Spotify pays from a pool by streamshare, so this is an average after the fact, not a rate.
4. The 1,000-stream floor. Spotify says "tracks must have reached at least 1,000 streams in the previous 12 months" to generate recorded royalties (Spotify for Artists); MusicTech describes this as demonetising tracks under 1,000 streams (MusicTech).
5. Recording side only, Spotify only, one year. We leave out publishing royalties: whether a prompt-generated song is a registrable composition is unsettled, and we do not assume it pays. Other platforms are left out too.

Stub 1: one track, one year

Break-even is $32.99 of costs divided by the rate: about 11,000 streams at $0.003, or about 6,600 at $0.005. That is our arithmetic from the cited prices and range.

Stub 2: the volume play (a year of output)

The obvious answer is to spread the costs. Use all 20 Pro downloads every month and the generator costs $8 ÷ 20 = $0.40 a track; one distributor plan covers the lot.

The two columns are 500 streams a track apart, and that gap is the difference between a loss and roughly a thousand dollars. The cost per track barely matters; the floor does.

What we think

• The floor is the whole game. Under the per-stream maths sits a binary: 999 streams earns nothing, 1,000 earns a few dollars. Any honest forecast for an AI track starts with "will this clear 1,000 in a year", not "what is the rate".
• Volume spreads costs, not reach. Uploading 240 tracks makes each one cheaper but does not give any of them listeners. A catalogue that each stalls at a few hundred streams is 240 zeros. In our view, the slot-machine model of AI uploads is mostly a way of paying generator and distributor subscriptions.
• Budget on the low rate. As in our million-streams stub, the $0.003–$0.005 range comes from distributor data and, we think, skews towards higher-paying markets. Bulk catalogues heard on free tiers or in lower-priced countries will likely land under it.
• Watch the downside, not just the upside. Spotify said it would start charging labels and distributors per track when it detects flagrant artificial streaming (Spotify for Artists). Anyone tempted to push tracks over the floor with bought plays is buying a penalty, not royalties. Distributor terms on AI content vary and change; check yours before uploading.
• Where it can work. A small number of AI-assisted tracks with a real audience, a playlist or a use (background music for your own videos, a game, a podcast) can clear the floor. The stub rewards finding listeners first and generating second.