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 V4 Flash is so popular that it's causing servers to crash. The official recommendation is for users to use another platform for now.

According to DynoBeat monitoring, after the official release of DeepSeek V4 Flash, it quickly experienced a surge in volume. The official API briefly returned a 503 error, indicating the service was overloaded, and advised users to temporarily switch to other large model API providers.

However, switching to a third party may not necessarily be cheaper. DeepSeek's official cache hit price is $0.02 per million tokens, while Alibaba Cloud, Tencent Cloud, and Volcano Ark under Byte all charge $0.2, which is 10 times more expensive.

The Agent will repeatedly read the code and context, with token caching often accounting for the vast majority. After moving to a platform with a higher cache hit price, the actual cost will increase significantly.

举报 Correction/Report
RelatedRecommended
Price is $3 per million Tokens for input, $6 per million Tokens for output, and $0.025 per million Tokens for cache hit input.
The road ahead no longer allows it to continue walking with its eyes closed.
DeepSeek V4-Flash Official Version API Launches Public Beta, Offering 1 Million Context, Pondering Mode, and Tool Invocation Capabilities at a Very Low Price.
Technology should never be an exclusive luxury.
DeepSeek Invests in Arborea IPO, Sparking Market Interest in Embodied AI Cost Reduction Pathway. The article points out that as the cost of machine cognition and action decreases simultaneously, AI may shift from a high-cost capability to a resilient infrastructure that can be iterated upon through trial and error.
Correction/Report
Submit
Add Library
Visible to myself only
Public
Save
Choose Library
Add Library
Cancel
Finish