Archive — CN — technode
DeepSeek to launch V4 in mid-July with new peak-time API pricing
DeepSeek is set to launch V4 in mid-July, featuring a 1-million-token context window and improved functionalities, alongside a new tiered pricing model introducing peak-time API charges.
Key points
- DeepSeek V4 scheduled for official release in mid-July.
- The new version enhances features and performance compared to the previous release.
- Standard context window size increased to 1 million tokens.
- Performance improvements include agent-based task execution, mathematical reasoning, and code generation.
- New pricing plan introduces peak (9 AM - 12 PM, 2 PM - 6 PM) and off-peak API usage rates.
Why it matters
- Peak-time pricing schemes are increasingly common as cloud services expand, reflecting broader IT trends in resource management.
- The focus on context window size and performance aligns with the growing demand for AI-driven solutions in enterprise applications.
- Similar enhancements in AI models are expected as competition increases among technology firms leading up to 2025.
DeepSeek V4's release illustrates the industry's shift toward flexible pricing and enhanced capabilities, marking a significant step forward in AI technology as competitors similarly adapt to evolving demands.