科大讯飞(002230) - 2024年5月20日投资者关系活动记录表
IFLYTEKIFLYTEK(SZ:002230)2024-05-20 15:48

Group 1: Project Progress and Development - The "Xunfei Super Brain 2030" plan has initiated seven projects, all of which are progressing steadily since the announcement in May 2022 [1] - The first four projects focus on core technology development, closely related to cognitive large models [1][2] - The "Xunfei Spark" large model has undergone five major upgrades since its release on May 6, 2023, continuously enhancing its seven core capabilities [2] Group 2: Technological Advancements and Achievements - The "Xunfei Spark" model has achieved leading positions in various international assessments, winning dozens of championships in AI competitions over recent years [2] - In the past three years, Xunfei has secured 17 international authoritative evaluation championships in the field of multimodal perception [2] - The AI virtual assistant for mental health screening has achieved a 95% consistency rate with specialist doctors, outperforming general practitioners by 34% [3] Group 3: Industry Applications and Impact - The "Xunfei Spark" model has been applied in various industries, including finance, automotive, energy, and education, with successful case studies emerging [4][5] - In the automotive sector, the first high-end new energy vehicle equipped with the "Xunfei Spark" model was launched in August 2023, receiving significant industry attention [4] - The AI virtual assistant for youth mental health has been implemented in over 3,000 schools, serving more than 2.5 million students [3] Group 4: Market Performance and User Engagement - The sales of Xunfei's smart hardware, including learning machines and translation devices, increased by 126% during the 2023 Double 11 shopping festival [6] - The "Xunfei Spark" app ranks first in download volume among general-purpose large model apps in China, with a high user satisfaction rating of 4.8 [7] - The user base of the "Xunfei Spark" app primarily consists of professional knowledge workers, with peak usage times during work hours [7]