- DeepSeek launched its flagship V4 Pro AI model, ending its preview phase
- V4-Pro-0813 offers improved AI agent capabilities for multi-step tasks
- DeepSeek to introduce peak and off-peak API pricing from August 17
Chinese artificial intelligence startup DeepSeek has officially launched its flagship V4 Pro model, moving the system out of its preview phase as the company steps up efforts to compete with a growing group of domestic AI rivals.
The newly released V4-Pro-0813 is available through DeepSeek's app, website and API.
The company said the latest version offers enhanced capabilities for AI agents, a key area as developers increasingly use AI systems to perform multi-step tasks rather than simply generate text or answer questions, according to Reuters.
The launch comes after an unusual development during testing. DeepSeek's lower-cost V4 Flash model outperformed an earlier preview version of V4 Pro in several independent evaluations, highlighting the rapid pace at which the company has been developing its AI models.
Also Read | AI Darkhorse: DeepSeek Weighs New Funding Just Weeks After $7-Billion Raise
Pricing strategy also changes
DeepSeek is also changing how it charges businesses using its models through APIs. The company plans to introduce separate peak and off-peak rates for V4 Pro and V4 Flash from Aug. 17.
Depending on the model and type of token usage, the price increases could range from 50% to more than 1,100%, marking a significant change in the pricing of DeepSeek's AI services.
The move comes as DeepSeek seeks to build on the global attention it received after its R1 reasoning model gained widespread attention in early 2025. Its rise challenged the assumption that advanced AI development necessarily requires spending on the scale of leading US technology companies.
Competition within China has also intensified. Companies including Moonshot AI, Zhipu AI, MiniMax, Alibaba and ByteDance have continued releasing increasingly capable AI models.
Also Read | China's DeepSeek Developing Own AI Chip To Cut Reliance On Nvidia: Report
DeepSeek expands beyond models
DeepSeek is now also scaling its business and infrastructure. The company is reportedly seeking nearly $8 billion in a new funding round at a potential valuation of around $74 billion, following approximately $7.4 billion raised in its first external financing round.
The startup plans to significantly expand its workforce, including teams focused on AI agents and data centres. It has also been hiring chip-design engineers as it explores developing its own AI processors, a move that could potentially reduce its dependence on suppliers such as Nvidia and Huawei.
Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.