Per token. Nothing else.

DeepSeek V4.1 Flash is sold at a peak rate and a cheaper off-peak one, because that is how the model vendor sells it. Peak is Mon-Fri 09:00-12:00 and 14:00-18:00 (UTC+8); every other hour, including the whole weekend, is off-peak.

Which rate you pay is decided when the request arrives, not when it finishes. A generation that starts at 17:59 and runs for four minutes is billed entirely at the peak rate, and one that starts a minute later is billed entirely at the off-peak rate. Nothing is split across the boundary, so a long request can never cost more than the rate you sent it at.

Every request records which window it was billed in, and it is on your own activity page — so a batch job moved into the cheap hours can be checked rather than taken on trust. Both rates and the window itself are on /v1/models as pricing, pricing_offpeak and pricing_window, so a client can price a request before making it.

You are billed per token against the usage the model vendor reports — the same numbers on your invoice line and in ours — with no rounding up to a minimum and no per-request fee.

Cached prompt tokens are billed separately, at the cache read rate rather than the input rate. Those already cost between a third and a twenty-fifth of a fresh input token depending on the model, and we pass that through rather than flattening it into one blended rate. On agent workloads with a long stable system prompt it is usually the largest saving of all.

Credit is reserved when a request starts and the unused part is refunded the moment it finishes, so a runaway generation stops at your balance instead of past it. Reasoning models are the exception worth knowing about: their reasoning is not bounded by max_tokens, so the reservation carries a measured allowance for it and an unusually long run can settle slightly above what was held.

A 4xx from upstream is never charged, and neither is a request that fails before any tokens are generated.