Meta Announces Muse Spark AI Model Led by Alexandr Wang

Meta Platforms Inc. announced Muse Spark, a new AI model developed by its Meta Superintelligence Labs. The model represents the company's first major large language model initiative since hiring Alexandr Wang, who leads the labs as chief AI officer.

Muse Spark targets advancements in multimodal reasoning and learning. Meta stated that Muse Spark's performance aligns with frontier AI models from competitors such as Google and OpenAI. The company has invested billions in AI development to narrow the gap with industry leaders.

Wang's leadership has been central to this effort.

Spark is integrated into Meta's own product ecosystem, including Meta AI.

It is not open source and lacks general availability through an API. Access remains limited to Meta's internal applications. Independent benchmarks indicate Muse Spark underperforms compared to models like Google's Gemini, Anthropic's Opus, and OpenAI's GPT variants.

Bloomberg reported that the model's capabilities do not yet surpass these rivals in key metrics. Meta has not disputed these assessments.

stock price increased by more than 8% in trading following the announcement.

The surge reflects investor optimism about the company's AI progress under Wang's direction. Trading volume rose significantly during the session. The announcement occurred amid Meta's broader push into AI, including prior investments in infrastructure and talent.

Wang joined Meta after leading Scale AI, bringing expertise in AI training data. No specific release date for wider deployment was provided. Competitors continue to dominate the AI landscape, with OpenAI and Google advancing rapidly in model capabilities.

Meta's strategy emphasizes internal use to enhance products like social media platforms and virtual reality tools. Analysts note the competitive pressures in the sector.