header-langage
简体中文
繁體中文
English
Tiếng Việt
한국어
日本語
ภาษาไทย
Türkçe
Scan to Download the APP
Partnered With:

DeepSeek even counts make-up workdays as off-peak, and today the API is still half-price.

Beating AI News Flash: DeepSeek supplements its API peak-valley billing rules. On weekends adjusted for work and Chinese statutory holidays, billing is at the off-peak rate all day. Although today, September 20, is a National Day adjusted workday, the API still charges the off-peak rate.


DeepSeek began implementing peak-valley pricing on August 17, with off-peak rates at half the peak rate. Less than a week later, it changed all Saturdays and Sundays to off-peak rates. This time, it further clarified the billing rules for adjusted workdays and statutory holidays.


Taking deepseek-flash as an example, for every million tokens of cache-miss input, the off-peak rate is 1 yuan and the peak rate is 2 yuan; for output, they are 4 yuan and 8 yuan respectively. October 10, another National Day adjusted workday, is a Saturday and is also billed at the off-peak rate.

举报 Correction/Report
RelatedRecommended
AI ends up in a showdown, competing to see who can turn the money-burning story into a profitable business first.
Technology should never be an exclusive luxury.
DeepSeek V4-Flash Official Version API Launches Public Beta, Offering 1 Million Context, Pondering Mode, and Tool Invocation Capabilities at a Very Low Price.
The success of DeepSeek has led to a panic in the U.S. capital markets. Share prices of companies such as NVIDIA, ARM, Broadcom, TSMC, Micron, AMD, Intel, and others have all experienced declines. Nasdaq 100 futures plummeted by -400 points, and over $1 trillion in market value has been wiped out from the U.S. stock market.
Price is $3 per million Tokens for input, $6 per million Tokens for output, and $0.025 per million Tokens for cache hit input.
Correction/Report
Submit
Add Library
Visible to myself only
Public
Save
Choose Library
Add Library
Cancel
Finish