Zhongtai: AI technology iteration and commercialization accelerate, highlighting the global competitiveness of domestic large-scale models.

Zhongtai released a research report stating that the global and Chinese large model market is growing rapidly, with a significant economic driving effect. Domestic AI is expected to build high-cost-effective token pricing barriers with the advantages of energy and computing power, continuously improving competitiveness. With the rising demand for AI to B/C applications such as AI Agent and multimodal, investment opportunities in companies with large model capabilities and application scenarios are promising. Zhongtai's key points are as follows: Technological breakthroughs are driving market growth, and the acceleration of commercialization highlights the investment value of AI Global AI technology achieved a breakthrough in 2025, with a significant improvement in the comprehensive capabilities of leading large models and multimodal understanding abilities, exceeding market expectations in terms of technological iteration speed. With the improvement of technology maturity and the release of payment willingness, the global large model market has entered a period of rapid growth, with a compound annual growth rate of nearly 81% in the next five years, indicating clear industry growth prospects. The scale of China's AI industry continues to expand, with core industry scale exceeding trillion yuan, and the number of enterprises and ecosystem are becoming increasingly perfect. The Chinese large model market also maintains high growth, with a projected compound annual growth rate of 58.5% from 2024 to 2026. Technological breakthroughs and industrial implementation resonate together, continuously highlighting the investment value of the AI race track. In 2025, multimodal technology continues to mature, and it remains the mainstream development trend in 2026. By 2025, multimodal large models will achieve deep integration, enhancing the capabilities of text, images, videos, and audio in a coordinated manner. Leading manufacturers have successively launched high-performance text-to-image and text-to-video models, with technology transitioning from the laboratory to commercial scale. The pace of multimodal iteration accelerates significantly throughout the year, expanding the quality of content, adaptability to scenarios, and integrity of the ecosystem. Looking into 2026, multimodal technology will continue to be the core development trend of AI. On one hand, it will continue to break through the boundaries of modes, enhancing complex reasoning and content creation capabilities; on the other hand, it will penetrate lightly into vertical industries, combined with copyright compliance and enhanced security systems, laying a more stable environment for large-scale commercialization. The scale of AI investment continues to expand, with energy and computing power building barriers for domestic competition in the international market Global IT investments in the AI field continue to grow rapidly, with strong long-term growth certainty, and AI significantly drives GDP growth. From the perspective of capital deepening efficiency, the penetration speed of AI is much faster than historical technological revolutions such as electricity and IT, and strong economic dynamics can be unleashed in the short term, making it a core investment theme. The essence of the AI industry is driven by computing power, energy, and capital, with electricity costs forming the core competitive advantage. China has significant advantages in electricity generation and prices, directly translating into advantages in domestic large model token pricing. In high-consumption scenarios such as AI Agent, cost barriers will further transform into a global competitive advantage. AI Agent further evolves, and the continuous expansion of domestic large model cost advantages Viewing AI Agent represented by OpenClaw, it has rapidly risen and has become one of the fastest-growing open-source projects on GitHub, capable of independently executing file processing, tool calls, cross-platform collaborations, and other complex tasks. At the same time, its token consumption has increased significantly. High-consumption scenarios further magnify the cost advantages of domestic models. China's energy and computing power support bases are sufficient and can meet the explosive demand for AI Agent. The cost-effectiveness advantage of tokens will continue to expand, providing a solid underlying support for the global expansion and commercialization realization of domestic AI. MiniMax: Building a full-mode model and application ecology MiniMax is a global AI company focused on full modes, using the MoE architecture and self-developed attention mechanisms to form a technological barrier in long text and multimodal fusion. The company adheres to the integrated route of "model + application," with rapid user growth and a complete model matrix covering language, video, speech, and music. The team has top AI backgrounds globally, diverse shareholder structures, high and continuous revenue growth, with over 70% of revenue from overseas. Although still in the investment period, user growth, payment conversion, and clear paths to commercialization are evident. The models rank among the top globally in various evaluations, showcasing outstanding technological competitiveness. Role model in going global: MiniMax's commercialization focuses on C-end AI native applications, with overseas markets becoming the main growth engine and significant effects in global layout. The product matrix is rich, covering general Agent, emotional companionship, video generation, speech music, and an open platform, with complete coverage of scenarios. Core products Talkie/ and MaxClaw have achieved successful landings, serving a large number of corporate clients through the open Claw ecosystem, with leading tool call volumes in the industry. With full-mode capabilities and cost advantages, the company has become a benchmark for the international expansion of domestic AI, demonstrating strong growth certainty. Risk warning: Risks include industry policy and regulatory risks, technological research and development falling short of expectations, intensifying market competition, profitability falling short of expectations, deviation in estimates, and delayed research report updates.