V-JEPA 2

Training compute

9.1×10²¹ FLOP

Parameters

1B

Published

Jun 11, 2025

V-JEPA 2 is an AI model developed by Facebook AI Research (United States and France), first published in June 2025. It works in the vision, video and robotics domain, on tasks such as robotic manipulation.

Training it took an estimated 9.1×10²¹ FLOP of compute (estimation method: operation counting). The model has 1,000,000,000 parameters.

Access: Open weights (unrestricted). Its weights are openly available. Epoch AI rates the confidence of this record as likely.

Full record

Organization: Facebook AI Research
Country of organization: United States, France
Domain: Vision, Video, Robotics
Task: Robotic manipulation
Training compute: 9.1×10²¹ FLOP
Compute estimation method: Operation counting
Parameters: 1,000,000,000
Model accessibility: Open weights (unrestricted)
Open weights: Yes
Epoch confidence: Likely