Meta unveils ‘Muse Spark,’ first AI model from Superintelligence team

Meta has unveiled “Muse Spark,” the first artificial intelligence (AI) model developed by its newly formed superintelligence team, marking a significant step in the company’s effort to catch up with industry leaders in the fast-moving AI race.

The announcement sent shares of the META higher, reflecting investor optimism that Meta’s heavy spending on AI could begin to yield results.

The new model is part of a broader initiative launched last year, when Meta made aggressive investments to rebuild its position in AI.

That push included hiring Alex Wang under a reported $14.3 billion deal and offering massive compensation packages to attract top engineering talent.

Muse Spark is the first release in a new internal series of models, codenamed “Avocado.” Unlike Meta’s earlier Llama models, which were widely distributed, the company is initially offering Muse Spark through a limited “private preview” and via its Meta AI app and website.

The model is designed to be lightweight and fast while still capable of handling complex reasoning tasks in areas such as science, mathematics and health.

In the coming weeks, Meta plans to integrate it across its ecosystem, replacing existing AI systems in apps like WhatsApp, Instagram and Facebook, as well as its smart glasses.

Independent benchmarks suggest Muse Spark is beginning to close the gap with top models from rivals such as Google, OpenAI and Anthropic in areas like language and visual understanding, though it still trails in coding and advanced reasoning tasks.

Mark Zuckerberg has previously cautioned that the first generation of models may not immediately lead the field, but emphasised the company’s long-term trajectory.

Meanwhile, Wang acknowledged the system still has “rough edges,” with larger and more capable versions already in development.

The launch also highlights Meta’s evolving business strategy for AI. The company is experimenting with integrating shopping features into its AI chatbot, allowing users to discover and purchase products directly within conversations.

With more than 3.5 billion users across its platforms, Meta is betting that embedding AI into everyday digital interactions will drive engagement and unlock new revenue streams — positioning the company to compete more aggressively in the global AI race.