Akamai Launches AI Grid Using Nvidia Blackwell Architecture

Akamai Technologies launched AI Grid on March 16, 2026, marking the first global-scale implementation of the NVIDIA AI Grid reference architecture. This orchestration system expands the Akamai Inference Cloud by routing AI workloads across more than 4,400 edge locations, regional sites, and core infrastructure to move intelligence away from centralized hubs.

Powered by Nvidia RTX PRO 6000 Blackwell Server Edition GPUs, the platform targets real-time applications requiring low latency, such as fraud detection for financial institutions, real-time dubbing for broadcasters, and gaming NPCs with sub-50-millisecond inference. The company aims to optimize what it calls tokenomics by reducing the need for round trips to centralized clusters.

Akamai validated the commercial scale of the project by announcing a four-year, $200 million service agreement with a major US technology company to deploy multi-thousand GPU clusters at the metro edge. While the infrastructure is operational, the company has not yet disclosed specific pricing, general customer availability, or the total number of integrated GPUs.