Investment Rating - The report assigns a strong investment rating to the industry, highlighting the leading position of the company in the Chinese large model sector and its potential for growth in the AGI space [2]. Core Insights - The company, founded in 2019, has established itself as a leader in the domestic large model industry, leveraging solid technical foundations and model capabilities. Its latest model, GLM-5, has achieved state-of-the-art (SOTA) performance in various benchmarks and received widespread acclaim from users [2][6]. - The commercial value of the models is being realized, with a projected revenue CAGR of over 130% from 2023 to 2027. API revenue is expected to be a major growth driver, with an estimated annual recurring revenue (ARR) related to APIs nearing 600 million yuan by early 2026, reflecting significant growth compared to the previous year [2][50]. - The AI Coding total addressable market (TAM) is estimated to reach a trillion yuan, with the company holding a competitive advantage in this area. AI Coding is identified as the fastest application scenario for AI, and the company is well-positioned to lead in this segment [2][94]. Summary by Sections Company Overview - The company has been a pioneer in the large model industry in China, releasing several significant models over the years, including GLM-10B and GLM-130B, and continuously iterating on its GLM series [6][12]. - The core value of the company lies in its GLM-based general large model capabilities, which have positioned it among the top tier of global large model providers [12][34]. Model Capabilities - The GLM series models have demonstrated exceptional performance in global benchmark tests, with GLM-5 ranking fourth globally in the Artificial Analysis Intelligence Index [24][25]. - The model matrix includes a wide range of applications, from language and multimodal models to intelligent agents, catering to diverse industry needs [28][30]. Business Model - The company's core business model is based on Model as a Service (MaaS), providing comprehensive model capabilities to users and facilitating the entire lifecycle from pre-training to deployment [36][42]. - The company offers differentiated MaaS deployment options, including cloud API calls, cloud privatization, and on-premises deployment, to meet various customer needs [42][44]. Growth Drivers - The company is experiencing rapid revenue growth, with a projected revenue of 310 million yuan in 2024 and a CAGR of over 130% from 2022 to 2024. The local deployment and cloud deployment revenues are expected to grow significantly [50][54]. - The rapid increase in token consumption indicates a strong demand for the company's models, supporting the anticipated growth in API-related revenue [54][55]. Industry Trends - The AI industry is witnessing accelerated growth driven by large models, with China's AI market projected to reach 1.607 trillion yuan by 2024, growing at a CAGR of 31% from 2022 to 2024 [67]. - The commercial path for enterprise-level large models is becoming clearer, with a significant portion of the market demand coming from enterprise clients [71][72].
中金 _ AI主线开年布局展望：智谱首次覆盖