Equinix, Inc. (Nasdaq: EQIX) announced a new distributed AI inference program called Equinix Inference Exchange, developed in collaboration with Nvidia and Together AI, according to a company press release.

The program combines Nvidia's Enterprise Reference Architectures, Together AI's inference platform — which supports more than 200 open-source models — and Equinix's global data center infrastructure. Connectivity between components will be provided through Equinix Fabric.

The announcement was made at Equinix Horizon, the company's inaugural customer and partner event, alongside a separate product called Equinix Fabric One.

Equinix Inference Exchange is intended to support enterprise AI inference workloads across distributed environments, including metro edge deployments, migration from proprietary to open-source models, and workloads subject to data residency requirements.

The solution will offer both multitenant and single-tenant deployment options. Together AI will operate the platform layer, while Equinix provides physical infrastructure including power, cooling, and network connectivity. Nvidia supplies the underlying compute architecture.

Equinix currently operates more than 280 data centers across 77 metros and lists over 10,500 businesses on its interconnection platform, along with 230 cloud on-ramps.

Equinix Inference Exchange is scheduled to become available in the first quarter of 2027.