Making advanced intelligence efficient, open, and deployable everywhere.

Fabric AI is a UK foundation-model research and deployment company building efficient AI models and infrastructure optimised for NVIDIA hardware.

We develop open and commercial foundation models designed to run across the computing spectrum—from edge devices and personal computers to professional workstations and data-centre accelerators.

Our goal is to make advanced AI faster, more accessible, more private, and easier to deploy on the hardware people and organisations already own.

What We Build

At Fabric AI, we focus on:

• Foundation Models for reasoning, coding, research, productivity, and general-purpose intelligence
• Efficient Small and Medium Models designed for local and on-device deployment
• NVIDIA-Optimised Models for Jetson edge platforms, GeForce RTX GPUs, NVIDIA RTX workstations, and NVIDIA data-centre accelerators
• Multimodal AI Systems capable of understanding text, images, and other forms of information
• AI Agents and Developer Tools for automation, software development, and real-world workflows
• Inference and Deployment Infrastructure for running models efficiently across edge, desktop, workstation, and data-centre hardware
• Open-Source Research and Tools that give developers greater control over how they build and deploy AI

Our Mission

Our mission is to make advanced intelligence efficient, open, and deployable everywhere.

We believe powerful AI should not be limited to a small number of closed platforms or expensive computing environments.

Fabric AI is building models and deployment technology that allow developers, researchers, businesses, and individuals to run capable AI systems locally, privately, and efficiently across the NVIDIA computing ecosystem.

Open and Commercial AI

Fabric AI combines an open-source ecosystem with commercial products and services.

We plan to release open-weight models, datasets, research, evaluation tools, and deployment software while also developing commercial models, hosted services, enterprise deployments, and custom AI systems.

This approach allows us to support open innovation while building sustainable products for real-world use.

Follow Fabric AI

This Hugging Face organisation will be used to publish Fabric AI models, datasets, demos, research experiments, benchmarks, and open-source projects.

We are building efficient foundation models for every class of NVIDIA computing—from edge devices to data centres.