Time-based pricing charges by the amount of time a product or service is used or accessed, such as per hour, per day, or per minute. Cost scales with duration rather than with units consumed.
Time-based pricing charges for how long something is used rather than how much of it. Per-hour, per-day, and per-minute rates are time-based: the meter is the clock. It is a natural fit when the cost or value of a service tracks duration, like compute time, equipment rental, or professional services.
How Time-based pricing works
The provider sets a rate per unit of time and charges for the duration of access or use. The clock starts when the customer begins (provisioning a server, renting equipment, starting a session) and stops when they end, with the bill equal to elapsed time times the rate. Some models round to billing increments (per second, per minute, per hour).
Time-based pricing is a form of usage-based pricing where the metered unit happens to be time. It works when duration correlates with cost (a running server consumes resources every second) or with value (an hour of expert help).
Time-based pricing examples
A cloud provider charges per second of compute while an instance runs. A GPU platform charges per GPU-hour. Equipment rental charges per day. A consultant charges per hour. A parking app charges per minute.
For AI and infrastructure, time-based pricing (compute-seconds, GPU-hours) sits alongside unit-based pricing (tokens, requests); products often use whichever better tracks their underlying cost.
Time-based pricing vs Unit-based usage pricing
| Time-based pricing | Unit-based usage pricing | |
|---|---|---|
| Meter | Duration (time) | Quantity (units) |
| Example | Per GPU-hour | Per token |
| Fits when cost tracks | Time running | Volume processed |
| Risk | Idle time still bills | Spikes in volume |
Benefits & when to use it
Time-based pricing fits services whose cost accrues with duration: running compute, rented hardware, live sessions, or expert time. It is intuitive for customers (a clear hourly or daily rate) and aligns revenue with the resource being held.
Its weakness is idle time: a customer billed per hour for a running server pays even when it does nothing useful, which can feel unfair and pushes customers to optimize duration. Where value tracks output rather than time, unit-based or outcome-based pricing fits better.
FAQ
What is time-based pricing?
Charging by the duration a product or service is used or accessed, such as per hour, day, or minute. Cost scales with elapsed time rather than with units consumed.
How is time-based pricing different from usage-based pricing?
Time-based pricing is usage-based pricing where the metered unit is time. General usage-based pricing can meter any unit (tokens, requests, GB); time-based specifically meters duration. Both scale the bill with consumption.
When does time-based pricing fit?
When cost or value tracks duration, like cloud compute, GPU time, equipment rental, or professional services. It fits less well when value depends on output rather than time held, where unit or outcome pricing is better.
How Credyt handles Time-based pricing
Credyt can meter time as the billable unit just as it meters tokens or requests. A wallet can hold compute-hours or seconds, and Credyt authorizes and debits elapsed time in real time against the balance, so a time-based product gets live balances, spend caps, and per-customer rate cards without building duration metering itself. Explore Credyt →