Never depend on breakage
Breakage — credits bought and never used — is real margin, and it's typically 15–40% of a consumer credit pack. But it's the first thing to disappear as your product gets stickier, and rollover policies or consumer-protection rules in some jurisdictions can eliminate it entirely. Build a pack that's profitable at 100% redemption and treat breakage as upside.
One credit should not equal one token
Abstract your credits away from raw model units. If a credit maps 1:1 to a token or an API call, every model price change forces a pricing change and customers can arbitrage your cheapest operation. Define a credit as a unit of value — one generation, one document, one minute — and internally price each operation in credits based on its actual cost.
- • Charge different credit amounts for cheap vs expensive operations.
- • Publish the credit cost of each action so usage feels predictable.
- • Give a monthly free allowance to keep light users engaged.
- • Re-price credits when model costs fall — or keep the price and take the margin.
Credits versus flat unlimited
Credits protect margin and make cost visible, but they add purchase friction and can suppress usage — users ration themselves, engage less, and churn. The common compromise is a subscription with generous included credits plus cheap overage: predictable for the buyer, capped for you.
Related guides
Long-form playbooks on the same topic, written by the RevenueLab editorial team.
LLM Token Costs in 2026: Pricing Every Model, Hidden Multipliers, and Margin Math
Input vs output token pricing across GPT, Claude, and Gemini, the context-window cost trap, how caching and batching cut bills 40–80%, and the real per-user margin most AI apps miss.
Read the guideSaaS Pricing Strategy: Per-Seat, Usage, Tiers, and the Hybrid Future
A framework for choosing a SaaS pricing model — when per-seat caps your growth, when usage-based makes revenue volatile, and how hybrid models stitch the two together.
Read the guideFAQ
How much should I charge per AI credit?
2–5× your raw cost per credit is the normal band. That covers payment fees, support, infrastructure, and leaves a software-like margin. Below 2× you have almost no room for the heavy tail.
What's a typical credit redemption rate?
60–85% for consumer packs. Enterprise prepay blocks run higher, 85–95%, because usage is planned. Always sanity-check your model at 100% redemption.
Should credits expire?
Expiry protects margin and forces re-purchase, but it generates support tickets, refund requests, and in some jurisdictions raises consumer-protection issues. A common middle ground is rollover up to a cap, e.g. two months' worth.
How do I convert tokens into credits?
Price each user-facing operation from its real token cost, then round into whole credits with a markup. One credit might be a short generation, three a long document. Keep the mapping stable even when underlying model costs change.
Are credits better than usage-based billing?
Credits are prepaid and cap your exposure while giving the customer certainty. Post-paid usage billing scales more naturally for enterprise but creates bill-shock risk and collections work. Consumer products lean credits; B2B leans metered.
How do I handle failed generations?
Refund the credit automatically. You still pay the model cost, so build a 5–15% failure buffer into your markup — charging for failures is the fastest way to generate refund requests and negative reviews.
How this calculator is built
Independently maintained
Written by Sam Doshi and the RevenueLab editorial team. We don't sell the data feeds this tool is built on.
Sourced from primary data
Benchmarks come from public AdSense / Stripe / IRS disclosures and reader-submitted data — never third-party "$X per view" claims. Full methodology.
Last editorial review
Reviewed on a rolling quarterly cycle. Dated reviews are published on the methodology record for each calculator.
Editorial standards
See our editorial policy and disclaimer. Results are estimates, not advice.