V-JEPA 2 is Meta's new world model, trained on video to understand and predict the physical world. It enables zero-shot robot planning and sets SOTA benchmarks in visual understanding. Model, code, and new benchmarks are now open.

A vision encoder setting new standards in image & video tasks. It excels in zero-shot classification & retrieval, surpassing existing models.

It is a hyper-realistic 10m Air Pistol Olympic Training VR experience designed for athletes, hobbyists, and curious first-timers. Whether you want to learn the basics, refine your technique or feel the thrill of standing in an Olympic arena this is your stage.

A state-of-the-art computer vision model trained with self-supervised learning (SSL) that produces powerful, high-resolution image features.

Now available with both 8B and 70B pretrained and instruction-tuned versions to support a wide range of applications

Chat with your choice of 28 AIs, now in beta. Each has a unique personality, mannerisms and backstory.

Some of Meta's AI characters are played by recognizable public figures.

Introducing the next generation of the Meta Training and Inference Accelerator (MTIA), the next in our family of custom-made silicon, designed for Meta’s AI workloads.