Free cloud & LLM pricing tools — AWS EC2 + LLM API Pricing Explorer

Off-Peak Pricing

Time-of-day pricing where the standard rate applies off-peak and a multiplier raises it during busy hours.

Off-peak pricing ties the per-token rate to the time of day. The provider publishes a standard, off-peak rate and applies a multiplier during peak hours when demand is highest. DeepSeek is the clearest example among the providers here: its listed prices are the off-peak/standard rates, and a 2× multiplier applies during peak windows of 09:00–12:00 and 14:00–18:00 Beijing time (UTC+8).

This rewards workloads you can shift in time. Batch jobs, overnight processing, and anything not tied to a user waiting in real time can run off-peak at the base rate; interactive traffic during business hours in the peak timezone pays the multiplier. Because the windows are defined in one timezone, whether "peak" lands during your users’ active hours depends on where they are.

This explorer shows the off-peak rate as the headline — the honest lower bound — and states the peak multiplier and window as a note, so you are never surprised by a bill that is double the sticker price. It is a genuine differentiator: off-peak pricing can make a model the cheapest option for schedulable work while being materially more expensive for real-time use in the wrong timezone. Confirm the current windows with the provider, as they can change.

FAQ

Which provider uses off-peak pricing?

DeepSeek. Its published rates are off-peak/standard, with a 2× multiplier during 09:00–12:00 and 14:00–18:00 Beijing time (UTC+8).

How do I benefit from off-peak rates?

Shift schedulable work — batch jobs, overnight processing — into the off-peak window so it runs at the base rate instead of the peak multiplier.

More LLM pricing terms

← Back to the LLM pricing hub