Skip to content
Brief

Meta Superintelligence Labs

Artificial IntelligenceDivision· part of Meta Platforms

Developing superintelligent AI systems that power Meta's advertising, consumer, and enterprise products while advancing open-source AI research for the broader ecosystem.

Last updated Sep 26, 2026 by ATDb automated enrichment · Connections updated Sep 28, 2026

Founded
2025
HQ
Menlo Park, California, United States
Connections
3

AI-drafted from public sources4 sources cited

At a glance

Employees
201-500
2integrations1corporate family

About

Meta's flagship AGI research lab competing directly with OpenAI, Google DeepMind, and Anthropic at the frontier of AI capability

Meta Superintelligence Labs (MSL) is Meta's advanced AI research organization, formed in 2025 as part of Meta's aggressive push toward artificial general intelligence (AGI). The lab was established by consolidating Meta's top AI talent and bringing in external leadership, most notably Alexandr Wang, founder and former CEO of Scale AI, who joined as Meta's Chief AI Officer to lead the initiative. The organization represents Meta's most ambitious bet on frontier AI development, operating with significant autonomy and resources within the broader Meta ecosystem. The lab focuses on foundational AI research and the development of next-generation large language models and multimodal AI systems, building on Meta's existing open-source AI efforts including the LLaMA model family. MSL is positioned to compete directly with OpenAI, Google DeepMind, and Anthropic in the race toward increasingly capable AI systems. Meta has committed billions in capital expenditure to support this initiative, including massive infrastructure investments in GPU clusters and data centers. From an AdTech and broader technology ecosystem perspective, Meta Superintelligence Labs is strategically significant because breakthroughs in AI capability directly feed into Meta's core advertising business — powering more sophisticated ad targeting, creative generation, audience modeling, and campaign optimization across Facebook, Instagram, WhatsApp, and Threads. The lab's work on foundation models also has implications for Meta's AI assistant products and its competitive positioning against other major platform players.

Business model

Internal R&D / Corporate Division

Target market

Enterprise

What they offer

  • LLaMA Models (Next-Gen)

    Next-generation large language models developed under MSL, continuing Meta's open-source LLaMA model family with significantly enhanced capabilities

  • Meta AI Assistant

    Consumer-facing AI assistant powered by MSL's frontier models, integrated across Meta's family of apps

  • Multimodal AI Systems

    Advanced AI systems capable of processing and generating text, images, video, and audio for use across Meta's platforms

  • AI Infrastructure Research

    Foundational research into AI training infrastructure, including custom silicon and distributed computing at massive scale

Key features

Frontier large language model developmentMultimodal AI research and deploymentOpen-source model releases via LLaMA familyIntegration with Meta's advertising and consumer platformsMassive compute infrastructure accessAGI-focused long-horizon research agenda

Use cases

Advanced ad targeting and audience modeling for Meta's advertising ecosystemAI-powered creative generation for advertisersConversational AI assistant across Meta appsContent moderation and safety systemsDeveloper tools and APIs for third-party AI applicationsEnterprise AI solutions built on Meta's foundation models

Customer segments

Meta's internal product teamsAdvertisers and marketers on Meta platformsDevelopers and researchers using open-source LLaMA modelsEnterprise customers adopting Meta AI tools

Tech & specs

Technology stack

PyTorch (Meta's open-source ML framework)LLaMA foundation modelsCustom MTIA AI accelerator chipsNVIDIA H100/H200 GPU clustersDistributed training infrastructureReinforcement Learning from Human Feedback (RLHF)Multimodal training pipelines

Security & compliance

Meta Platform Security StandardsGDPRCCPA

Deployment

Cloud

API

Limited

Corporate history
  1. 2025 · Founded
Connection details

Explore further

2 views