How they use Nebius AI Cloud
Cerebrium uses Nebius compute within its AI deployment infrastructure. Its documentation identifies Nebius as the provider for us-central1 and eu-north1.
Applications can run globally or be pinned to a region. Cerebrium’s operational status history also connects Nebius to its inference infrastructure. It records a Nebius network incident under the EU Inference API on September 2, 2026, and a Nebius cluster incident under both US and EU Inference APIs on October 15, 2026.
Cerebrium marked the October 15 incident resolved at 17:51 UTC, with US and EU clusters back online.
Sources
3 sourcesNebius Clusters are unavail... | Cerebrium
Published 15 Oct 2026 · status.cerebrium.ai · Web page
Cerebrium reported that Nebius was down in the US and EU on October 15, 2026 at 17:34 UTC. At 17:51 UTC, it reported that the Nebius issue was resolved and its US and EU clusters were back online.
Open original sourceCerebrium operational status history
Listed 2 Nov 2026 · Cerebrium · Web page
Cerebrium’s EU Inference API history records a Nebius network incident on September 2, 2026, and a Nebius cluster incident on October 15, 2026. The US Inference API history also records the October cluster incident.
Open original sourceCompany context
cerebrium.aiCerebrium provides serverless infrastructure for engineering teams building and operating AI applications. Its platform runs voice agents, large language models, image and video models, training tasks, and batch processing workloads on GPU and CPU resources. Developers can deploy their own code or containers, with automatic scaling, regional deployment, storage, and monitoring handled through the platform. Customers include Tavus, Deepgram, and ResembleAI. The service supports streaming and WebSocket endpoints for interactive applications and charges according to compute usage.