Western tech stocks, including all-conquering Nasdaq-listed Nvidia Corp, tumbled ahead of opening today as hype grew over the weekend about Chinese tech firm DeepSeek’s latest AI model, which claims to be cost-effective while running on less-advanced chips.
DeepSeek, which was only founded in 2023, claims that its latest model is able to utilise reinforcement learning without the need for human supervised fine-tuning.
The Chinese advances cast doubt over the stratospheric valuations for companies like Nvidia, which has led the global AI stock boom as its chips have been essential to the technology. Shares in the Santa Clara, California-based firm slid 10% in premarket trading on Monday.
READ MORE: Who is behind DeepSeek? Chinese startup redefining AI and rattling global markets
Nasdaq 100 futures also fell 3.4%, while contracts on the S&P 500 fell 2% as of 5 a.m. in New York. In Europe, tech stocks led market losses, with shares of chip equipment maker ASML Holding NV down 11%.
Northern tech firms in the AI space may be interested to learn that, unlike many leading AI developers, DeepSeek has committed to make its models and training protocols entirely open-source. Models like DeepSeek-V3 and DeepSeek-R1 are publicly available, allowing anyone with the right tools to test, validate, and build upon their work.
To support the research community, DeepSeek has so far released a suite of open-source models, including DeepSeek-R1-Zero, DeepSeek-R1, and six dense models with parameter sizes ranging from 1.5B to 70B. These models, distilled from DeepSeek-R1, are built on widely recognized architectures such as Qwen and Llama, ensuring compatibility with existing tools and workflows.
