Riyadh - Sharikat Mubasher: OmniOps, the Saudi sovereign AI infrastructure technology company, announced the launch of the Bunyan Inference Hub to help governments, enterprises, startups, and developers access Saudi-hosted open-source AI models and computing resources.
The engine enables organizations to keep their data and AI workloads within the Kingdom, access a catalogue of AI models or bring their own, and use computing resources without managing separate infrastructure for each model or use case, OminOps revealed in a press release today.
Customers access predefined token bundles across supported models and workloads, with dedicated GPU infrastructure available for organizations that need additional capacity or isolation.
The new engine is enabled by OmniOps’ proprietary inference optimization engine and developed by the company’s R&D team. It adapts how AI models run on available hardware to match workload requirements, helping organizations improve GPU utilization, reduce cost per token, and maintain consistent performance as demand grows, without having to manage the underlying infrastructure.
Mohammed Altassan, founding CEO of OmniOps, said: “The new hub brings together our proprietary Bunyan Inference technology and experience in sovereign AI deployment in a simplified platform that public and private sector organizations can plug into their operations and use to deploy AI solutions more efficiently.”
The launch of Bunyan Inference Hub comes as demand for AI inference infrastructure grows globally, with more organizations moving AI models into production.
In July 2025, OmniOps launched Bunyan, the Kingdom's first sovereign Inference as a Service platform, to deliver cutting-edge AI capabilities with full support for text, vision, and speech applications while ensuring data sovereignty and compliance. Two months later, the company partnered with NourNet to develop and operate Bunyan to provide enterprises with end-to-end, sophisticated AI solutions that are compliant with local data sovereignty regulations.