en.Wedoany.com Reported - CITIC Securities' latest research report indicates that since 2026, domestic large model manufacturers have focused on upgrading Agent and coding capabilities, competing to release new models. We believe the upcoming next-generation DeepSeek model is expected to continue the high-cost-effectiveness open-source route, achieving stronger memory functions and ultra-long context processing capabilities. While refining its code and Agent abilities, it will also address the multimodal shortcomings, bringing new investment opportunities in the areas of model original manufacturers, AI applications, and AI infrastructure. This analysis from CITIC Securities once again directs market attention to the release expectations of DeepSeek's next-generation model and the potential industrial chain changes it may trigger.
From an industry background perspective, competition among domestic large models has entered a new stage since 2026. Major manufacturers are no longer solely pursuing parameter scale expansion but are shifting their focus to refining practical application capabilities such as artificial intelligence agents and code generation. CITIC Securities believes that under this trend, DeepSeek's next-generation model continuing the high-cost-effectiveness open-source route holds significant strategic value. This route has previously helped DeepSeek build a broad user base within the developer community. If the new model can maintain this advantage while further strengthening memory functions and ultra-long context processing capabilities, it is expected to develop differentiated competitiveness in complex task processing and large-scale information analysis scenarios.
CITIC Securities' report particularly emphasizes that while refining code and Agent capabilities, DeepSeek's next-generation model will also address its multimodal shortcomings. This means the new model will not only continue to optimize in traditional text processing fields but is also likely to possess understanding and generation capabilities for multimodal content such as images and audio. Addressing multimodal capabilities will expand the application scenarios of DeepSeek's next-generation model from conversational assistants and programming aids to broader fields such as visual content analysis and multimedia creation, thereby opening up incremental space for commercial implementation.
From an investment perspective, CITIC Securities points out that the release of DeepSeek's next-generation model will bring new investment opportunities in three directions: model original manufacturers, AI applications, and AI infrastructure. For model original manufacturers, large model companies with core technical capabilities will benefit from technological iteration and ecosystem expansion. At the AI application level, enterprises developing applications based on high-performance open-source models are expected to lower R&D barriers and accelerate product implementation. In the AI infrastructure field, related industrial chain segments such as computing power, data, and model services will also experience demand growth alongside the enhancement of model capabilities. CITIC Securities believes that if DeepSeek's next-generation model continues the high-cost-effectiveness open-source route as expected, it will further validate the feasible path of open-source models within the commercial ecosystem, providing an important reference for the technological evolution and industrial collaboration of domestic large models.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
