Documentation/Getting Started/Overview
Overview
vibezcheck is the simplest way to add Stripe billing and token metering to any AI app. You write 1 line of code, and vibezcheck tracks prompt tokens, calculates dollar costs in real time, and sends usage to Stripe with zero added latency.
What is VibezCheck?
When you build an AI app using models like GPT-4o, Claude 3.7, or Gemini, AI providers charge you for every token your users generate.
If you want to charge your users for their AI usage or give them a monthly credit budget, you normally have to:
1. Set up a database and table schemas to log every token.
2. Build custom webhook listeners and background queues.
3. Buffer AI streams in intermediate proxies (which slows down response time).
vibezcheck replaces all of that with a single line of code.
How It Works (The Mobile Data Analogy)
Think of AI tokens like mobile phone data (gigabytes):
- ✓Every time your user sends a prompt, they consume a small amount of data (input tokens).
- ✓When the AI writes a response, it generates output data (output tokens).
- ✓vibezcheck acts like a digital speedometer: it measures the exact token count, converts it into real dollars (e.g. $0.0015), and syncs it with your Stripe account automatically.
Why You Need It
- ✓⚡ Zero Latency Overhead: Streams pass directly from OpenAI/Anthropic to the user's browser with 0ms buffering delay.
- ✓🧠 Reasoning Token Aware: Accurately tracks hidden thinking tokens in GPT-5, o3-mini, and Claude 3.7 that traditional counters miss.
- ✓🛡️ Runaway Loop Protection: Built-in circuit breakers halt runaway AI loops before they drain your bank account.
- ✓💳 Instant Monetization: Automatically provisions Stripe customers and sends metered billing events without writing backend cron jobs.
- ✓🆓 100% Free in Local Development: Works out of the box with zero Stripe keys required for local testing.
💡 Tip: Use Left / Right arrow keys to navigate guidesPage 1 of 11