BuilderPrice
New DeepSeek peak/off-peak pricing

DeepSeek API Pricing Calculator

Estimate DeepSeek V4 Flash and V4 Pro token costs with the new peak and off-peak API pricing. Enter cache-hit input tokens, cache-miss input tokens, output tokens, and a UTC call hour to see the cost difference.

DeepSeek API cost calculator

Estimate DeepSeek V4 Flash / V4 Pro costs from cache-hit input, cache-miss input, and output tokens.

Official pricing

Selected window

Off-peak

Selected-hour cost

$1.6720

Peak cost

$3.3440

Off-peak savings

$1.6720

$1.6720 off-peak cost · 50% off-peak discount
Local time reference: 20:00 GMT+8

Peak hours are 01:00-04:00 and 06:00-10:00 UTC; all other hours are off-peak.

Pricing effective: 2026-08-16 16:00 UTC. DeepSeek pricing can change; verify the official pricing page before subscribing or batching production calls.

Peak windows

Peak hours are 01:00-04:00 UTC and 06:00-10:00 UTC. The start time is included and the end time is excluded, so 04:00 UTC is already off-peak.

Off-peak batching

Move non-urgent jobs such as batch summarization, offline extraction, evals, and embeddings-like preprocessing into off-peak windows when your product can tolerate delayed work.

Cache matters

DeepSeek prices cache-hit input, cache-miss input, and output tokens separately. High cache reuse can materially lower your effective cost.

FAQ

What are DeepSeek API peak hours?

DeepSeek lists peak hours as 01:00-04:00 UTC and 06:00-10:00 UTC. All other hours are off-peak. The new peak/off-peak pricing takes effect at 16:00 UTC on August 16, 2026.

How much cheaper is DeepSeek off-peak pricing?

DeepSeek states that off-peak rates are half of peak rates. This calculator shows both costs so you can estimate savings before batching work.

Why separate cache hit, cache miss, and output tokens?

DeepSeek prices cache-hit input, cache-miss input, and output tokens separately. A workload with high cache hits can be much cheaper than one that repeatedly sends new context.

Is this calculator official?

No. BuilderPrice is an independent calculator based on public DeepSeek pricing docs. Always verify the official DeepSeek pricing page before production usage.

When should I batch calls into off-peak windows?

Batch non-urgent evaluation, indexing, synthetic data, summarization, and offline processing into off-peak windows when latency is not critical. Keep interactive user-facing calls on the schedule that fits your product.