A version of Chinese startup DeepSeek's flagship AI model is by far the least expensive to run on benchmark tests among well-known models globally, according to a research firm. The model is more than 100 times cheaper to run than Anthropic's Claude Fable 5, a leading model in the industry. This revelation positions DeepSeek as a formidable player in the AI landscape, particularly as it prepares for a potential IPO in early 2025.

The research firm Artificial Analysis, based in San Francisco, conducted the analysis and found that DeepSeek's V4-Flash charges $0.14 per million input tokens and $0.28 per million output tokens. These figures are significantly lower than those of competing models, including Moonshot AI's Kimi K3, which costs 86 cents per test, and OpenAI's GPT-5.6 Sol, which costs $1.86 per test. This disparity in cost highlights the efficiency of DeepSeek's model in comparison to others.

According to the research, DeepSeek's V4-Flash has an average cost of 3 cents per test, which is far below the costs of other models. This efficiency is attributed to the model's advanced architecture and optimization techniques. The findings suggest that DeepSeek's model could be a game-changer for businesses looking to reduce their AI operational costs without compromising on performance.

The implications of this development are significant for the AI industry. Companies that rely heavily on AI models for their operations may find DeepSeek's model to be a more cost-effective solution. However, the lower cost may also lead to increased competition, potentially affecting the market dynamics. Additionally, the efficiency of the model could influence the strategies of other AI providers, prompting them to enhance their offerings or adjust their pricing models.

Despite the potential advantages, the model's performance and scalability must be closely monitored. As the AI landscape continues to evolve, the long-term impact of DeepSeek's model on the industry remains to be seen. The research firm's findings are a starting point, and further studies will be necessary to fully understand the implications of this development.