Meta AI
Profile Overview
Meta AI is the dedicated artificial intelligence research division of Meta (formerly Facebook), established in 2013 under the leadership of Yann LeCun. Operating under the umbrella of Fundamental AI Research (FAIR), Meta AI is characterized by its open weights philosophy, which prioritizes releasing research models to the developer community to accelerate global innovation and set open standards. The division's most prominent release is the Llama (Large Language Model Meta AI) series, including Llama 3 and Llama 3.1, which have become the industry standard for open weights foundation models. By providing accessible, high-performance weights, Meta AI has democratized AI research and enabled developers and companies worldwide to deploy custom, fine-tuned models without relying on closed APIs. Beyond language modeling, Meta AI has pioneered computer vision research with models like Segment Anything (SAM) and developed PyTorch, the leading open-source deep learning framework. Meta AI integration plays a vital role across Meta's consumer suite, powering the Meta AI assistant across Instagram, WhatsApp, Messenger, and Facebook. Because it operates as an internal research division funded directly by Meta's massive corporate advertising revenue, Meta AI does not raise external venture capital. However, its technology contributes directly to Meta's hardware products (such as Ray-Ban Meta smart glasses) and the company's broader social platform ecosystem.
Last Financing Round
Flagship Offerings
Vetted AI Models
Llama 4 Maverick
activeMeta's next-generation open weights model. Delivers premium agentic capabilities, reasoning, and tool call compliance for local or self-hosted enterprise stacks.
Llama 4 Scout
activeMeta's ultra-long context model for document processing and massive retrieval tasks.
Llama 3.3 70B Instruct
activeMeta's state-of-the-art open weights model, providing enterprise-grade reasoning and logic. Exceptionally powerful for self-hosted customer support, text generation, and tooling workflows.
Llama 3.2 11B Vision
activeMeta's lightweight open weights vision model, optimized for mobile devices and local deployments. Capable of visual understanding, chart reading, and fast text generation.