Multimodal generative AI systems
Multimodal generative AI systems can accept multiple types of inputs, such as text and images, and produce various forms of output.
AI systems that generate text
Meta's text-generating AI systems can answer questions and have conversations based on text entered as prompts.
AI systems that generate images
Meta’s image-generating AI systems can create images based on descriptive text entered as prompts.
Instagram feed ranking
Feed ranking for Instagram prioritizes content to help people see the posts they’re most likely to find interesting, or want to engage with.
Social commerce model is an AI system that helps connect people to products from businesses and creators that are interesting to them.
Our approach to explaining ranking
AI systems inform the ranking of content for many experiences on Meta's products, such as viewing Facebook Feed, watching reels on Instagram or browsing Facebook Marketplace.
Our mission is to ensure that AI at Meta benefits people and society. System cards are one of the ways we deliver on this intention through their role in creating more transparency and control around how a person’s data is collected and used.Learn more about how we’re designing AI responsibly
Models and libraries
Our open-sourced libraries and models for those taking our AI learnings further through software and app development
Large-scale datasets and benchmarks for training, evaluating, and testing models to measure and advance AI progress
Our demos for anyone wanting to experience our latest research breakthroughs first hand
Multiple machine learning (ML) models that help people understand the intention, impact and limitations of our AI systems
Our library of published papers to learn about our latest AI breakthroughs and innovations