Reka

Reka is a research lab, staffed by more than 50 AI researchers and engineers from Google DeepMind and Meta including creators of Gemini and Veo, building multimodal AI models that process and generate video, image, audio, and text for what it calls the physical AI era, supporting applications like robotics and wearable devices. It offers native multimodal processing in a single model architecture with real-time inference, video processing infrastructure for tagging, reasoning, searching, and clipping large datasets, training data infrastructure called Claru for egocentric video and robotics trajectories, and enterprise API and infrastructure called Infer. Pricing is not publicly disclosed; Reka directs interested parties to request a demo for enterprise engagement.