Meta

Advanced AI model for interactive object segmentation in media.

Tools for creative expression using generative AI technology.

Generative AI experiences for Meta Products worldwide.

Meta AI can analyze and answer questions about photos.

Expands feature to explain AI-driven content recommendations.

AI models provide second opinion on content before enforcement.

Advanced AI assistant built using Llama open-source model.

Labels for AI-generated content across audio, image, video.

AI features improve business chat interactions on WhatsApp.

Advanced AI assistant available for free across Meta apps.

Text-only model with reduced serving costs.

High-performance vision-language encoder with curated datasets.

Evaluation toolbox for text-to-image generative models.

Method for scaling memory layers to enhance factuality.

Tokenizer-free model with dynamic patching for byte-level processing.

New training paradigm for hierarchical language modeling.

Program-guided data generation for theory of mind reasoning.

Generative paradigm for multiple modalities including images and audio.

Open source model for neural video watermarking and traceability.

Foundation model for controlling virtual embodied agents' behavior.

sEMG dataset for AI hand pose estimation models.

sEMG-based typing dataset for AI model development.

Analyzes suspected deepfake content in audio or video form.

AI provides election information with authoritative source links.

Self-supervised model for multilingual speech processing tasks.