Meta announced that its newly formed Superintelligence Labs has produced its first internal AI models, a milestone that follows six months of development. Chief Technology Officer Andrew Bosworth described the models—codenamed Avocado for text and Mango for image and video—as “very good” and noted they are still being refined before they can be deployed to Meta’s consumer and enterprise products.
The achievement comes amid a $70‑$72 billion AI infrastructure buildout that Meta began in 2025. The company is spending roughly $18.8 billion each quarter on data‑center expansion, including a 2 GW facility that will house more than 1.3 million Nvidia GPUs. These investments are intended to give Meta a competitive edge over Google, Microsoft, and OpenAI by enabling faster, larger‑scale model training and inference.
Superintelligence Labs was created in June 2025 to accelerate Meta’s move from research to production. The lab’s focus on high‑impact models is supported by a talent‑war strategy that has attracted researchers such as Alexandr Wang and Nat Friedman, with compensation packages reportedly reaching $100 million. The new models are expected to power Meta’s next‑generation ad‑targeting algorithms, recommendation engines, and the Meta AI assistant, potentially unlocking new revenue streams and improving user engagement across Facebook, Instagram, and other platforms.
Bosworth emphasized that the models are “six months into the work” and that the lab is still refining them. He added that the progress addresses investor concerns about the pace of Meta’s AI investments and signals that the company is moving beyond experimentation toward operational tools that can be monetized. Zuckerberg echoed this sentiment, noting that the lab’s success marks the beginning of a new era for Meta’s AI ambitions and that the company is committed to leading the industry in AI‑driven products.
The development of these internal models also reflects Meta’s broader strategy to shift from an open‑source philosophy toward proprietary, high‑performance models that can be tightly integrated into its ecosystem. By controlling the entire AI stack—from data centers to model training to product deployment—Meta aims to reduce reliance on external cloud providers and to capture a larger share of the AI‑enabled advertising market.
The milestone is a tangible sign that Meta’s aggressive AI spending is starting to yield results. It positions the company to better compete with rivals that have already released large language models and to capitalize on the growing demand for AI‑enhanced content curation and advertising. The continued refinement of Avocado and Mango will likely accelerate Meta’s ability to monetize AI across its platforms, while the substantial infrastructure investment underscores the company’s long‑term commitment to AI leadership.
The content on EveryTicker is for informational purposes only and should not be construed as financial or investment advice. We are not financial advisors. Consult with a qualified professional before making any investment decisions. Any actions you take based on information from this site are solely at your own risk.