share_log

科大讯飞(002230):讯飞星火V4.0全面对标GPT-4TURBO AI行业应用再次升级

iFLYTEK (002230): iFLYX Starfire V4.0 fully targets GPT-4TURBO AI industry applications once again

浙商證券 ·  Jun 28

Key points of investment

On June 27, the company officially released the iFLY Spark Model V4.0 and related industry applications. While improving the model's seven core capabilities, the GPT-4 Turbo was fully benchmarked. On the basis of the iteration of the big model, the company's applications in education, office, medical care and other fields have once again achieved major iterative upgrades.

The core capabilities of Spark V4.0 have been fully upgraded. GPT-4O, a leading company in financial and medical applications, officially released the Spark Big Model V4.0 based on the country's first 10,000 card computing power cluster “Flying Star 1”. The core capabilities are fully benchmarked against GPT-4 Turbo, and further upgraded in terms of graphic recognition capabilities, leading GPT-4O in application effectiveness in scientific research, finance, medical care, and administration of justice. In the live display, Starfire V4.0 excelled in complex instructions, complex logical reasoning, and spatial reasoning.

The company released the Spark smart reader, and the AI+ education application matrix continues to improve its ability to recognize complex graphics in educational scenarios based on the Spark V4.0 model. The company released the first Spark smart reader, which can achieve functions such as intelligent correction, academic analysis, and personalized work, which is expected to greatly improve teachers' efficiency. At the same time, the iFLY AI learning machine equipped with the Spark model has been further upgraded to the AI 1-on-1 question-answering and tutoring function, which can perform interactive exploratory learning, provide heuristic explanations to students through multi-modal abilities, and the company's AI+ education application matrix continues to improve.

The iFLYX App has been significantly upgraded, and the core capabilities of the iFLYX StarFire Model V4.0 in the medical field have all surpassed GPT-4 Turbo. On this basis, the various capabilities of the iFLYX App have been continuously upgraded, covering 1,600 common diseases and 2,800 common drugs. Up to now, the total number of downloads of the iFLYX App has reached 12 million. The launch of the “Personal Digital Health Space” of the iFeixiao Medical App can build a personal digital health space based on personalized user data such as electronic medical records, examination reports, and medical examination reports. The Spark Smart Platform was officially released. The AI Agent is expected to accelerate penetration into various B-side segment scenarios. The Spark Enterprise Intelligence Platform released by the company can cover 400+ AI atomic capabilities, integrate 90+ external signal sources, and open up 100+ internal IT systems. It can be used by enterprises to quickly build intelligent applications and platforms that can be implemented in conjunction with business scenarios 32 smart devices have been launched around production, science and innovation, office scenarios, etc., and enterprise users can plug and play. Based on the enterprise intelligence platform, the company has created typical application cases such as Spark Business Opportunity Assistant and Spark Bidding Assistant to sample enterprise applications.

Profit forecasting and valuation

We believe that as the company continues to iterate in the field of AI large model technology and the continuous improvement of the AI+ industry application business ecosystem, the overall operation is expected to grow. We expect the company to achieve revenue of 243.10/293.80/ 35.022 billion yuan in 2024-2026, with year-on-year growth rates of 23.71%/20.86%/19.20%, corresponding net profit to mother of 12.99/20.15/2,520 billion yuan, respectively, with year-on-year growth rates of 97.63%/55.13%/25.04%, corresponding EPS of 0.56/0.87/1.09 yuan, maintaining the “buy” rating.

Risk warning

Risks brought about by increased market competition; risks brought about by order growth or implementation falling short of expectations; risks brought about by AIGC related technology implementation and commercialization not falling short of expectations;

The translation is provided by third-party software.


The above content is for informational or educational purposes only and does not constitute any investment advice related to Futu. Although we strive to ensure the truthfulness, accuracy, and originality of all such content, we cannot guarantee it.
    Write a comment