Liquid AI is an artificial intelligence company spun out of MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL). The firm is focused on reimagining foundation models from first principles, developing what it calls Liquid Foundation Models - designed to operate efficiently across a range of hardware, from large-scale data centres to resource-constrained edge devices. Its approach moves beyond traditional transformer architectures, pursuing adaptive AI systems built on alternative technical foundations.
The company's technical work spans foundation models, efficient and adaptive AI, edge deployment, and transformer alternatives. This positions it at the intersection of cutting-edge AI research and practical, hardware-aware model design - an area of growing interest as demand increases for AI that can run outside cloud environments.
Liquid AI operates with a flat, meritocratic culture that emphasises scientific rigour and autonomous execution. The organisation describes itself as a no-playbooks environment where assumptions are challenged and teams rebuild from fundamentals. Transparency, early problem escalation, and individual ownership are cited as core working principles, alongside a stated commitment to aligning company objectives with employees' personal and professional development.