Nebius closes Eigen AI deal, adds inference optimization

Nebius says it has completed its acquisition of Eigen AI. The company said the deal closed on June 10, after it was announced on May 1.

What’s new in the mix is Eigen AI’s focus. Nebius describes Eigen AI’s technology as centered on inference and model optimization—basically the part of the AI pipeline that turns trained models into outputs in real-world use, and tries to do that more efficiently.

Nebius is positioning the acquisition as another step toward a broader AI platform. The company says the deal supports its plan to build a full-stack AI platform that covers work from training through to production deployment.

For people who just want AI tools to work reliably and not drag down systems, inference optimization is the practical piece. Even when training is done elsewhere, the day-to-day cost and speed of running models can make a big difference. So adding more expertise aimed at inference may help Nebius improve how its cloud services handle models once they’re actually being used.

Nebius also framed the move as part of strengthening its AI cloud stack. That matters because many customers experience AI performance in production, not in a demo environment—latency, throughput, and efficiency tend to be the issues that show up first.

The acquisition itself is just one corporate milestone, but it signals where Nebius expects its future platform capabilities to land: better model optimization for inference, integrated into a platform that spans the path from development to deployment.

Source: Nebius