Nunchux AI
Building the Frontier of Multimodal Inference
About Nunchux AI
Nunchux AI builds infrastructure that makes multimodal generative AI faster and cheaper to serve. The company develops a multimodal inference acceleration engine and serving platform for image, video, and world models, with optimization across NVIDIA GPUs, AMD GPUs, and mobile devices.
Industry
Artificial Intelligence & Machine Learning
Headquarters
San Francisco, CA
Founded
2026
What we do
Multimodal inference optimizationImage generation servingVideo generation servingModel quantization and compressionAPI infrastructure for generative AIGPU acceleration (NVIDIA, AMD)
Interested in Nunchux AI?
Join our talent pool to be considered for future opportunities.