DeepSeek Launches V4 Pro and V4 Flash AI Models with Huawei Chip Support and Competitive Pricing
Agentic AI at $0.14 per million tokens with Chinese chip support...
DeepSeek has released two new open-source AI models: V4 Pro and V4 Flash. Both models showcase significant upgrades in agentic capabilities—meaning they can autonomously plan and execute multi-step tasks—as well as enhanced knowledge recall and logical reasoning. The V4 Flash variant is priced aggressively at $0.14 per million input tokens, making it one of the most cost-effective options for developers building agent-driven applications.
A key differentiator is hardware support: these models are partially powered by Huawei computer chips, marking a strategic move to reduce reliance on US-based chipmakers like NVIDIA. This could reshape supply chain dynamics and open new avenues for AI deployment in regions with restricted access to American semiconductors. For enterprises, the combination of low cost and alternative hardware compatibility makes DeepSeek's latest release a compelling option for scaling AI workloads.
- V4 Flash costs $0.14 per million input tokens, highly competitive for agentic AI tasks
- Models support Huawei chips, reducing dependence on US semiconductor suppliers
- Improved agentic capabilities allow autonomous multi-step task execution
Why It Matters
Affordable, open-source AI models with Chinese chip support could disrupt pricing and hardware dependencies globally.