Core components
1. Products
Organize your services. Each product is a folder for something you sell - an LLM API, data processing pipeline, or analytics platform.2. Items (optional)
Name the SKUs you sell, so charges can be mapped to your accounting system. An item is thin — a name, the product it belongs to, optional metadata, and external references to your CRM, ERP, or tax provider. Tagging a billable metric or fee with an item is optional and changes nothing about how it bills.3. Features
Gate access and enforce quotas. Attach features to prices so customers automatically receive the right entitlements when they subscribe. Use boolean flags, static config values (e.g., seat limits), or metered quotas that decrement as customers consume.4. Billable metrics
Track consumption. Whether it’s tokens processed, queries run, or gigabytes stored - if you can measure it, you can bill for it.5. Fees
Add fixed charges. Setup fees, monthly subscriptions, platform access - any predictable cost that doesn’t vary with usage.6. Plans
Package your pricing. Offer different tiers to different customer segments - from hobbyists to enterprises.7. Prices
Define what each unit costs. For example, charge a flat rate per unit, take a percentage of the metered value as a revenue share, or set a volume price so the rate falls as usage grows.How it works
Start with a product, then add metrics to track usage and define fees for fixed charges. Create plans for different customers, attach prices to each metric within those plans, and link features to prices to control what each tier can access. Each product can have multiple plans, and each plan combines fees with metric prices and feature entitlements. This lets you offer the same service at different price points to different customer segments. Each price charges in advance, at the start of the period, or in arrears, at the period end. One plan can mix both. See Payment terms. Items sit outside this sequence. Create them when you need your invoice lines to carry an external SKU, GL, or tax code — typically when connecting an accounting system — then tag the relevant metrics and fees with them. Existing pricing keeps working untouched until you do.Pricing flexibility
A plan sets the default price for each customer segment. Credits, limits and contract terms let you change what one customer gets and pays without a new plan.Credits that track usage in real time
A metered entitlement gives each customer an allowance of credits every period. The balance reflects consumption in real time, so your app can check it before each request. Use a hard limit to stop usage at zero, or a soft limit to allow overage. Unused credits can roll over, with a cap or a floor. See Metered entitlements.Prepaid and promotional credits
Sell credits in advance through a grant purchase. The customer pays through hosted checkout, and Paygentic creates the grant when the payment completes. Add free credits directly for promotions or support adjustments. A grant can have an expiry date, and you can void it. See Grants.Contract terms on one subscription
Bill a contracted minimum, or cap the billed quantity, with a quantity bound. Change a price or a quantity from a given date on one subscription with price intervals. For a negotiated deal, create a customer-specific plan.Price changes without migrations
Publish a new plan version to change prices. Each subscription either moves to the new version at its next billing period, or keeps its current prices. A period that is already invoiced never changes.Real-world example: LLM provider
Let’s say you’re launching an LLM service: Product: “Neural Language Engine” Metrics to track:- Input tokens consumed
- Output tokens generated
- Fine-tuning compute hours
- Model API calls
- Input: $0.002/1K tokens
- Output: $0.006/1K tokens
- 10K free tokens monthly, as a metered entitlement
- Input: $0.001/1K tokens, half the Developer rate
- Output: $0.003/1K tokens
- Fine-tuning: $3/hour
- Priority queue access, as a boolean feature
- Negotiated rates on a customer-specific plan
- A contracted minimum of 10M input tokens a month, billed whether or not the customer uses them
- SLA guarantees
Data platform example
Running a data warehouse service: Product: “Cloud Analytics” Metrics:- Query compute seconds
- Storage GB-months
- Data egress GB
- Concurrent connections
- On-demand: A standard price per compute second, billed in arrears
- Volume: A volume price on storage, so the rate per GB falls once the period’s usage passes 100 GB, and again past 500 GB
- Prepaid: Customers buy blocks of compute seconds in advance through a grant purchase
- Contracted minimum: An annual deal that commits the customer to 10 TB of storage a month, billed whether or not it is used