IFLYTEK(SZ:002230) Tianfeng Securities·2024-10-25 00:00Investment Rating - The investment rating for the company is "Buy" with a maintained rating [1]. Core Views - The company has turned profitable in Q3 2024, marking the beginning of a new growth trajectory driven by "large models + embodied intelligence" [1]. - The revenue for the first three quarters of 2024 reached 14.85 billion yuan, a year-on-year increase of 17.73%, while the net loss was 340 million yuan [1]. - The company has focused on key strategic business areas, ensuring operational health and accelerating cash collection [1]. Financial Performance - In Q3 2024, the company reported a single-quarter revenue of 5.525 billion yuan, up 15.77% year-on-year, with a net profit of 56.96 million yuan, a significant increase of 120.87% year-on-year [1]. - The non-net profit for Q3 was 14.87 million yuan, reflecting a year-on-year growth of 173.64% [1]. - As of the end of Q3 2024, accounts receivable stood at 14.003 billion yuan, an increase of 17.29% compared to the beginning of the year [1]. Business Segments - The large model has shown significant effects, driving the transformation and upgrading of traditional businesses [1]. - In the education sector, revenue from Q1 to Q3 2024 grew by 22%, with sales of learning machines increasing by over 100% [1]. - The medical business saw the cumulative downloads of the "Xunfei Xiaoyi" app reach 12 million, with a user satisfaction rate of 98.8% [1]. - Consumer products experienced a 70% year-on-year increase in sales during the "618" shopping festival [1]. Future Outlook - The company has adjusted its revenue forecasts for 2024-2025 from 30.42 billion yuan and 38.163 billion yuan to 22.858 billion yuan and 26.701 billion yuan, respectively [1]. - The net profit forecast for 2024-2025 has been revised down to 577 million yuan and 886 million yuan, with a new forecast for 2026 of 31.09 billion yuan in revenue and 1.131 billion yuan in net profit [1]. - The company is positioned as a "national team" in artificial intelligence, with expectations for non-linear acceleration in commercialization of large models [1].