Your AI app might be losing money on its heaviest users. Here's how to check.
Description and use
FiscaLLM lets subscription‑based AI apps track per‑user costs, identify loss‑making users, and set alerts and caps.
Problem
Subscription‑based AI apps can incur higher costs on heavy users than the revenue they generate.
Outcome
The tool allows developers to identify users whose AI usage costs exceed their subscription revenue, enabling targeted cost controls and alerts.
Workflow
- 1. Add a backend between the app and the AI provider to intercept calls.
- 2. Use the FiscaLLM SDK to report usage after each AI call, passing userId and feature.
- 3. Connect revenue data via a RevenueCat or Stripe webhook, or send payments through the revenue API.
- 4. View margin per user, plan, and feature in the dashboard.
- 5. Configure budget alerts, runaway‑loop alerts, and optional caps with meter.check(userId).
Evidence
I built FiscaLLM to answer one question: which users cost me more than they pay?
The SDK reads only the model name and token counts. It never sees prompts or answers.
Paste a RevenueCat or Stripe webhook URL, or send payments yourself with the revenue API. FiscaLLM joins revenue and cost on the user ID.
- Record date: 2026-10-23T01:11:12.020Z