AI cost calculator
DeepSeek API pricing calculator
DeepSeek charges half price outside peak hours (01:00 to 04:00 and 06:00 to 10:00 UTC on weekdays). Enter your workload and how much of it runs at peak to see the monthly bill on V4 Pro and V4.1 Flash.
Your inputs
Result
$606.94 a month on DeepSeek V4 Pro
$2.02 per 1,000 requests, with 21% of traffic in peak hours.
- Cost per request
- $0.0020
- Cost per 1,000 requests
- $2.02
- Per month
- $606.94
- Per year
- $7,283
- Saved by prompt caching
- $231.59 a monthagainst the same requests with no cache hits
- Cheapest DeepSeek model for this workload
- DeepSeek V4.1 Flash$164.44 a month
Worked examples
Each row is the calculator run on the inputs shown, with every other input at its default.
| Workload | Model | Requests a day | Tokens in / out | Peak share | Per month |
|---|---|---|---|---|---|
| Support chatbot, around the clock | DeepSeek V4.1 Flash | 5,000 | 3,000 / 300 | 21% | $224.01 |
| Same chatbot on Flash | DeepSeek V4.1 Flash | 5,000 | 3,000 / 300 | 21% | $58.32 |
| Batch job run at night UTC | DeepSeek V4.1 Flash | 20,000 | 8,000 / 600 | 0% | $2,962 |
| Asia business hours | DeepSeek V4.1 Flash | 10,000 | 4,000 / 500 | 60% | $1,130 |
| Coding agent, all peak | DeepSeek V4.1 Flash | 2,000 | 60,000 / 2,000 | 100% | $1,093 |
How it is calculated
DeepSeek prices input on a cache hit, input on a cache miss and output separately, and every price has a peak and an off-peak rate. Off-peak is exactly half of peak.
Peak hours are 01:00 to 04:00 and 06:00 to 10:00 UTC, Monday to Friday, except Chinese public holidays. That is 35 of the 168 hours in a week, so traffic spread evenly runs about 21% at peak. Weekends and Chinese public holidays are off-peak all day.
The model name deepseek-flash is served by DeepSeek-V4.1-Flash-1017; the older names deepseek-v4-flash are billed at the Flash price. DeepSeek's Search and Fetch APIs cost $0.75 per 1,000 requests and are not included here.
Questions
How much does the DeepSeek API cost?
Per million tokens at peak: V4 Pro is $1.32 input (cache miss), $0.044 cached and $3.96 output; V4.1 Flash is $0.30, $0.006 and $1.20. Off-peak prices are half.
When are DeepSeek off-peak hours?
Every hour except 01:00 to 04:00 and 06:00 to 10:00 UTC on weekdays. Weekends and Chinese public holidays are off-peak all day.
How much does DeepSeek V4 cost compared with Claude or GPT?
For the same tokens, V4 Pro costs a fraction of Claude Sonnet 5.5 or GPT-6.1 Sol. The grey bars in the chart show those models for your workload at their standard prices.
Does cached input really cost that little?
Yes. A cache hit on V4 Pro is $0.044 per million tokens at peak, about 3% of the cache-miss price, so long reused prompts are cheap.
Sources
- DeepSeek: Models and pricing, checked 2026-10-29
Prices and plans change. This page was last checked on 2026-10-29. The data is open: llm-prices.json (CC BY 4.0).