#Gate广场四月发帖挑战 #Meta推出AI模型MuseSpark On April 8, local time, Meta officially launched its new AI model Muse Spark. Its stock price surged nearly 9% in a single day, and ultimately closed up 6.5%, increasing its market value by more than 100 billion USD. This is an important milestone following Meta’s restructuring of its AI strategy, marking its return to the core circle of the AI race.

Technical breakthrough: Muse Spark’s comprehensive evaluation score is 52 points, placing it among the world’s top four, with standout multimodal capabilities (chart reasoning score of 86.4 points, ranking first) and leading performance in health Q&A (42.8 points). It adopts “thinking compression” technology, reducing computing power requirements to 1/10 of the previous generation, and supports unified processing of text, images, and speech.
Strategic transformation: It gives up the open-source route and shifts to closed-source commercialization, planning to monetize through API services and a subscription-based model. The model is deeply integrated into platforms such as Facebook and Instagram, covering 3 billion daily active users, and it rolls out scenario-based applications such as shopping assistants.
Market response: On the day of the release, Meta’s stock price hit an intraday high of nearly 9% and closed up 6.5%. The capital market recognizes its ability to make up for its shortfalls in AI commercialization and to rebuild its technology stack, believing it has formed a new “four-pole competition” landscape alongside OpenAI, Google, and others.
Challenges and shortcomings: Coding ability (terminal coding 59 points) and abstract reasoning (42.5 points) still lag behind leading models, and it needs to address controversies in the developer ecosystem caused by the closed-source approach as well as the pressure from rapid iteration by major tech companies.