DeepSeek V4 Official Release Set for Mid-July with Peak-Valley API Pricing
Claire Weston
DeepSeek announced its V4 production release for mid-July, introducing a peak-valley pricing model for API calls — peak-hour rates double the standard price, forcing enterprise users to rethink their calling patterns.
What is actually changing in V4?
DeepSeek confirmed that V4's production release will bring feature upgrades and performance improvements.
No specifics have been disclosed yet. The "production release" label signals that the model has cleared internal validation and is moving from testing to stable deployment.
How does peak-valley pricing work?
The new mechanism splits API usage into peak and off-peak tiers: peak price = 2× the standard rate.
Peak hours are 9:00 AM – 12:00 PM and 2:00 PM – 6:00 PM daily — 7 hours total.
This means → only about 30% of the day is billed at the higher rate; the rest stays at the current price. In plain terms = shift your calls outside those windows and your cost doesn't change.
What happens to enterprise bills?
For companies that concentrate API calls during peak hours, effective costs will rise significantly.
Whether they can absorb the increase by shifting calls off-peak becomes the key variable in procurement decisions.
This means → peak-valley pricing is essentially a choice: change when you call, or accept a higher bill. Small developers are barely affected; high-volume enterprise users feel it most.
Why is DeepSeek doing this?
The official reasoning: "more efficient resource allocation and better service stability."
In plain terms = peak-hour request volume is overwhelming the servers. The price signal redirects some traffic to off-peak windows, easing system load.
This reflects that DeepSeek's API traffic has grown to the point where active congestion management is necessary — a sign that monetisation is maturing, even if it comes with growing pains.
Content is for reference only, not financial advice.