News
U.S. companies "abandon" head AI: Chinese models break through with high cost performance Recently, a subtle trend has emerged in the global AI application market
2 min read
Source: Telegram AI频道
U.S. companies “abandon” head AI: Chinese models break through with high cost performance Recently, a subtle trend has emerged in the global AI application market. As many leading American AI manufacturers continue to adjust their pricing strategies, many American companies have begun to proactively seek more cost-effective alternatives, and large Chinese models are accelerating their entry into the application links of American companies with their excellent cost performance. According to the latest industry data, since February 8 this year, the proportion of American companies calling mainstream Chinese models such as DeepSeek and Zhipu GLM has increased significantly. According to statistics from the AI model aggregation platform OpenRouter, the proportion of calls made by these domestic models on a weekly basis remains above 30%, with the peak reaching 46%. Compared with the data of the past year, this growth trend is very rapid. Industry analysts believe that behind this change is the company's strict consideration of AI cost control. When the task does not require the so-called "strongest" model, enterprises are more likely to route to those "good enough and cheap" models. According to relevant data calculations, domestic open source and open weight models have shown strong competitiveness in terms of cost. Compared with the top cutting-edge models in the United States, the price is generally 60% to 90% lower. In addition to price advantages, catching up in technical strength is also the key to the breakthrough of domestic models. Currently, the performance gap between domestic head models and top American cutting-edge models has been shortened to 6 to 9 months. In real-world application benchmark tests, some Chinese models perform within a hair's breadth of performance on specific tasks compared to leading U.S. models, yet cost only one-fifth of the latter. Many start-ups have begun to switch models to reduce costs and increase efficiency. Industry cases show that by migrating workflows from high-priced models to cost-effective domestic models, companies can save millions of dollars in operating expenses within months. This performance-based and cost-driven competitive strategy is allowing Chinese AI models to show unprecedented vitality in the international application market. via AI News (author: AI Base)