Building the Fastest Generative Media Infra

Nunchux AI is building the next frontier of multimodal inference. We’re a San Francisco team making image, video, and world models up to 100× cheaper and faster at the same fidelity, so real-time generative media can run anywhere, from data centers to phones. Our team brings together veterans from AMD, Snowflake, and NVIDIA, and researchers from MIT, CMU, and Berkeley behind key innovations on efficient generative models. We’re inventing the infrastructure powering the next generation of multimodal GenAI, and your work will reach millions of users and developers. If that excites you, we’d love to talk.

Reach out at [email protected]. Open applications are welcome, and we read each one.