头部模型官宣降价，应用端或加速落地

Investment Rating - The industry investment rating is maintained as "Leading the Market - A" [1] Core Viewpoints - Major model providers have announced significant price reductions, which are expected to expand the market and lower user costs for both B-end and C-end applications, accelerating the explosion of AI applications [1] - The price of the API input for Alibaba Cloud's Qwen-Long model has dropped by 97% to 0.0005 yuan per thousand tokens, making it approximately 1/400 of the GPT-4 price [1] - The report suggests that the rapid price cuts by leading model companies are likely to drive the application end's growth and further reduce content costs, empowering both B-end smart industries and C-end creators [1] Summary by Sections Investment Highlights - On May 21, Alibaba Cloud announced a significant price cut for its Qwen-Long model, while Baidu made its two main models free of charge [1] - The price of the pro-32k inference input for ByteDance's Volcano Engine model is 0.0008 yuan per thousand tokens, which is 99.3% cheaper than the industry average [1] - The report anticipates that public cloud and API will become the mainstream way for enterprises to use large models, with expected exponential growth in API call volumes [1] Market Performance - The industry is projected to perform 20% better than the CSI 300 index in the coming year [1] - The report indicates a 0% change in the market performance for the current period, with a forecasted decline of 20% in the next quarter [1] Analyst Insights - The report emphasizes that the ability to adapt, perform, and price effectively will determine which algorithm teams will dominate the edge AI field in the coming years [1] - The rapid advancements in domestic large model teams are highlighted, with examples of companies like OpenAI and Google leading in parameter scale [1]