Recently, the esteemed investment fund and consulting firm Andreessen Horowitz unveiled the highly anticipated 2025 Global Top 100 Generative AI Consumer Applications List. This list, segmented into web and mobile versions, each featuring 50 generative AI applications, aims to evaluate the global popularity and technical prowess of these AI applications.

Among them, Chinese AI enterprise DeepSeek shines brightly, with its AI application securing the second spot on the web version, trailing only industry giant ChatGPT.

Since its official launch on January 20, 2025, DeepSeek has swiftly amassed a vast user base due to its advancements in deep learning and neural network technology, particularly its exceptional performance in image generation and natural language processing. Within just ten days of its debut, DeepSeek soared to the second position on the global AI product ranking, trailing only ChatGPT, underscoring its robust technical capabilities and user appeal.

Its mobile product, launched on January 25, reached the 14th spot on the monthly active user ranking within five days and surged to second place in February. Despite user numbers dipping due to policy-related bans in some countries, its overall performance remains impressive.

DeepSeek's success is not a fluke. Following the official release of the DeepSeek-R1 large model early in the year, the company has been relentless in enhancing user experience through technological innovation. DeepSeek's AI chatbot swiftly garnered significant traffic post-launch, achieving rapid user growth in a short span. Statistics reveal that 21% of DeepSeek's users hail from China, 9% from the United States, and 8% from India, evidencing its global user base.

In addition to DeepSeek, other notable Chinese large models also excelled on this list. ByteDance's Doubao ranked 10th on the web version, Dark Side of the Moon came in 11th, Hailuo Video from large model unicorn MiniMax placed 12th, and Kuaishou Keling secured the 20th spot, all surpassing previously popular products in the same field like Sora, Midjourney, and Runway. On the mobile ranking, Baidu AI Search ranked 4th, Quark AI 6th, Doubao 7th, and Meitu 17th, further showcasing the formidable strength of Chinese AI enterprises.

DeepSeek's accolade is not just a testament to its technical prowess but also a boost to the global influence of Chinese AI enterprises. With continuous technological advancements and expanding application scenarios, DeepSeek is poised to continue spearheading development trends in the AI field, offering more intelligent and convenient services to users worldwide.

Multiple Chinese Products Among the World's Best

The global investment fund Andreessen Horowitz has released the 2025 Generative AI Consumer Application List, with several Chinese products performing admirably and ranking highly. Notably, Hangzhou-based DeepSeek emerged as this year's biggest dark horse, with its mobile user growth surpassing ChatGPT's record.

Moreover, in the realm of AI large models, China has also achieved remarkable feats. The latest list released in February 2025 by ChatbotArena, a globally renowned AI model evaluation platform, shows that Chinese AI large models occupy multiple prominent positions.

DeepSeek's DeepSeek-V3 and DeepSeek-R1 have garnered significant industry recognition for their outstanding performance and cost-effective, high-efficiency development strategy. Additionally, Alibaba Cloud's Tongyi Qianwen series of large models performed impressively, with Qwen2.5-Max ranking seventh globally with a score of 1332, demonstrating robust technical strength.

The breakthroughs of Chinese AI large models are not just limited to technology and performance but also extend to their innovative development models and open-source contributions. Taking Alibaba Cloud as an example, its Tongyi large model family comprehensively covers language, image, video, audio, and other modalities, with performance ranking among the world's top tier.

Furthermore, Alibaba Cloud actively promotes the construction of an open-source ecosystem. The Tongyi model has been continuously open-sourced and has become the most popular domestic large model among enterprises and developers. The number of derivative models in the Qwen series has also reached 74,300, surpassing Llama for the first time to become the world's largest family of generative language models.

The collective breakthrough of Chinese AI large models has sparked deep reflections within the industry. Traditionally, it was widely believed in the AI field that only massive investments could lead to the development of high-performance large models. However, Chinese enterprises are rewriting this narrative through innovative low-cost development models. They have not only achieved technological breakthroughs but more importantly, discovered a sustainable development path.

The significance of this development model transcends the technical level. It significantly lowers the entry barriers for AI technology, granting more enterprises the opportunity to participate in large model development. Simultaneously, cost reduction also translates to more competitive prices for users, fostering the popularization and application of AI technology.

With the ascendancy of Chinese AI large models, the global AI competition landscape is undergoing profound transformations. The success of Chinese AI enterprises not only showcases China's technological innovation strength and potential but also brings new development opportunities and challenges to the global AI field. The remarkable performance of Chinese AI large models globally is not just a strong testament to China's technological innovation strength but also an important reshaping of the future global AI competition landscape.

Booming AI Large Model Industry

With the acceleration of digital transformation and the growing demand for intelligence, China's AI large model industry is embracing unprecedented development opportunities. Recently, a series of developments related to the AI large model industry have underscored China's robust momentum and vast prospects in the field of artificial intelligence.

According to the latest data, the market size of China's AI large model industry reached approximately 29.416 billion yuan in 2024 and is projected to surpass 70 billion yuan by 2026. The continuous expansion of this market size not only mirrors the strong demand for AI large model applications among Chinese enterprises but also indicates the immense development potential of the industry.

Concurrently, with continuous technological advancements and expanding application scenarios, AI large models have demonstrated their potent enabling role in various applications such as entertainment, e-commerce, work, and vertical fields, enhancing business efficiency and intelligence levels.

At the local government level, Suzhou City is at the forefront of promoting the development of the AI large model industry. On February 14, 2025, Suzhou hosted the "AI+" Innovation Development Promotion Conference, where several measures for high-level construction of the "AI+" Innovation Development Pilot Zone were announced, along with 14 policies supporting core technology research and the innovative application of large models, with individual support reaching up to 100 million yuan. The introduction of these policies has undoubtedly injected robust momentum into the development of the AI large model industry in Suzhou.

Moreover, Suzhou has gathered numerous core enterprises, constructing a complete enterprise gradient system ranging from science and innovation-oriented small and medium-sized enterprises to industry leaders, initially forming a relatively complete industrial chain encompassing the basic layer, framework layer, model layer, and application layer of artificial intelligence.

In terms of technological innovation, Chinese enterprises have also made significant strides. For instance, the development of multimodal large models enables AI to more comprehensively understand and process complex scenarios, enhancing intelligence levels. Additionally, the application of new technologies such as the Self-playRL (self-play reinforcement learning) paradigm has further optimized the technical route and performance of large models.

Furthermore, domestic AI tools like DeepSeek have effectively reduced dependence on high-end chips through technological breakthroughs and ecological openness, demonstrating extensive application potential and practical value across multiple industries and fields.

In terms of application scenarios and implementation, AI large models have achieved remarkable outcomes in various fields such as finance, healthcare, education, and manufacturing. For instance, in the healthcare sector, the ability of AI large models to assist in diagnosis and disease prediction has significantly improved; in education, the capability of AI large models for personalized learning and intelligent tutoring has been widely applied; and in manufacturing, the application of AI large models has promoted the automation and intelligent optimization of production processes.

China's AI large model industry is poised to embrace broader development prospects. As technology further matures and application scenarios continue to expand, AI large models will play a pivotal role in more fields, driving the intelligent transformation and upgrading of various industries. Simultaneously, the government will continue to increase investment and support for AI technology, providing a favorable policy environment for the growth of the AI large model industry.