A Chinese startup specializing in artificial intelligence, DeepSeek, has recently shared data on the expenses and revenues associated with its popular models V3 and R1. The startup claims a theoretical profitability ratio of up to 545% per day, although it cautions that actual earnings will be significantly lower. This marks the first time a company from Hangzhou has disclosed financial data for the post-training phase – when trained AI models perform various functions, for instance through chatbots.
DeepSeek emphasizes that the disclosed data highlights the remarkable efficiency of its models. The focus is on the less resource-intensive “inference” stage rather than the production-heavy training process. The exceptionally high profitability ratio suggests substantial investment potential in AI technologies, achieved with minimal initial capital outlay. Moreover, this transparency comes at a time when global skepticism towards AI stocks has increased following a sharp decline in January, after widespread adoption of chatbot applications built on V3 and R1 models.
The release of these figures has generated significant attention against the backdrop of declining investor confidence in AI-related stocks beyond China. The surge in popularity of chatbot applications earlier this year led to widespread market fluctuations, rendering the disclosure of deep data on financial efficiency particularly noteworthy. While the theoretical profitability is impressive, real-world results are expected to be tempered by operational complexities and market dynamics.
Assessment of Fundamental Indicators
1. Equipment Costs – The company spent less than US$6 million on chips used for training, a figure considerably lower than competitors like OpenAI (ticker: OPEN).
2. Operational Efficiency – High performance during the inference stage is key to achieving notable profitability ratios.
3. Growth Forecasts – Sustaining these indicators could have a positive impact on market confidence in AI technology stocks.
- Innovative data processing methods
- Reduced training costs
- High adaptability of solutions across various applications
- Open data disclosure for independent evaluation
DeepSeek’s detailed financial disclosure underscores the increasing importance of data transparency and operational efficiency in the competitive landscape of artificial intelligence. By detailing its expenditure on crucial components like training chips, the startup demonstrates a cost-effective approach compared to American counterparts. Although theoretical figures highlight an impressive profitability ratio, the real-world earnings are naturally expected to be lower, reflecting the complexities inherent in applying advanced AI solutions in practical environments.
This announcement may further influence market perceptions regarding tech companies outside China, particularly at a time when investors are critically evaluating asset performance in the AI sector.
1 Comments
The acquisition under discussion could signal a transformative moment, introducing innovative dynamics into the automation software sector