Introducing DINOv3: Self-supervised learning for vision at unprecedented scale

Open Molecules 2025 and Meta’s Universal Model for Atoms: Design at the Atomic Scale | AI at Meta

Unlocking How the Human Brain Develops Language | AI at Meta

Adjoint Sampling: A Breakthrough in Highly Scalable, Reward-Driven Generative Modeling | AI at Meta

New Research: Molecular Property Prediction, Language Processing, and Neuroscience | AI at Meta

Built with Llama: Solo Tech | AI at Meta

Meta PARTNR: Unlocking Human-Robot Collaboration

Advancing robotics and touch perception | AI Research from Meta FAIR

Knowledge Distillation with Llama 3.1 405B | Llama for Developers

Open Approach to Trust & Safety: Llama Guard 3, Prompt Guard & More | Llama for Developers

Meet Llama 3.1

Understanding the Llama 3 Tokenizer | Llama for Developers

More ways to run Llama 3 | Build with Meta Llama

Run Llama 3 on Linux | Build with Meta Llama

Run Llama 3 on Windows | Build with Meta Llama

Run Llama 3 on Mac | Build with Meta Llama

10 Years of Advancing the State-of-the-Art through Open Science at FAIR | AI at Meta

Reimagining Meta's Infrastructure for the AI Age | AI at Meta

MSVP - Meta's First In-House Silicon for Video Processing | AI at Meta

MTIA - Meta's First-Generation AI Inference Accelerator | AI at Meta

Accelerating Research with Meta's Research SuperCluster | AI at Meta

Origin Stories | AI at Meta

Inside the Lab: Building for the metaverse with AI (2022) | AI at Meta

CICERO from the research team’s POV | AI at Meta

CICERO as a Diplomacy partner | Meta AI

Speaking of CICERO with Mike Lewis | Meta AI

Diplomacy Gameplay with CICERO | AI at Meta

CICERO & Strategic Reasoning with Noam Brown and Anton Bakhtin | Meta AI

Milestones to CICERO | Meta AI

CICERO: The first AI to play Diplomacy at a human level | AI at Meta

Is human led mathematics over? Panel with Joelle Pineau, Timothy Gowers & Yann LeCun | Meta AI

The first AI-powered speech translation system for a primarily oral language | AI at Meta

Make-A-Scene: Mike Schroepfer talks with Oran Gafni about text and sketch prompts

NLLB: Schrep talks with Necip Fazil Ayan about inclusion through the power of AI

Make-A-Scene: Greater creative control for AI image generation.

200 languages within a single AI model: A breakthrough in high-quality machine translation

Acoustic synthesis for AR/VR experiences

Studying the brain to build AI that processes language as people do

AI for clean energy: Mike Schroepfer talks to Larry Zitnick about Open Catalyst

Using AI to deliver more inclusive biographical content on Wikipedia

Project CAIRaoke: Breakthrough conversational AI that makes talking to assistants more enjoyable

Teaching AI to translate 100s of spoken and written languages in real time

Builder Bot demo

Introducing the AI Research SuperCluster — Meta’s cutting-edge supercomputer for AI research

Harmful content can evolve rapidly and it’s crucial for AI systems to evolve alongside it.

Building diverse MiniHack environments with just a few lines of code

Habitat 2.0: Training home assistants to rearrange their habitat

Zero-shot transfer to Formula 1 racetracks

Ego4D: Teaching AI to perceive the world through your eyes

AI now enables robots to adapt rapidly to changing real world conditions

PyTorchVideo: A deep learning library for video understanding

Expire-Span: Teaching AI How to Forget at Scale

Advancing the state of the art in computer vision with self-supervised Vision Transformers

How can we make sure that AI systems operate fairly for everyone?

How can we teach AI skills using visual demos? Model-based Inverse Reinforcement Learning

Mike Schroepfer @Schrep chats "AI 101 with Joelle Pineau" co-lead of the Facebook AI Research lab.

How does AI learn to finish your sentences? Learn about Smart Compose

How AI is getting better at detecting hate speech

The first AI model that translates 100 languages without relying on English data

FastMRI breakthrough shows AI-accelerated MRIs interchangeable with traditional MRIs