# Highlights - June 2026

• Added an FAQ section with product-specific pages for IaaS, Speedboat, AI Platform, AI Model Service, and Key Management. See FAQ.
• Expanded the Speedboat documentation with product overview, included components, deployment characteristics, and typical use cases.
• Introduced Hosted Cluster, a new IaaS offering: a dedicated OpenShift Kubernetes cluster with its own control plane on Phoeniqs sovereign infrastructure. See the Hosted Cluster guide for what's included, pricing, licensing, and getting started.
• Promoted inference-glm-51-754b (GLM 5.1) from sandbox to Active Models after successful performance, stability, and cost evaluation; the model is now generally available on the Phoeniqs Model Service with production pricing and throughput.
• Removed GLM 5.1 from Sandbox Models.
• Scheduled inference-glm-51-754b (GLM 5.1) for decommissioning; see Scheduled for Retirement. Suggested replacement: inference-glm-52-754b (GLM 5.2).
• Added inference-glm-52-754b (GLM 5.2) to Sandbox Models for performance, stability, and cost validation; promotion to the active models catalog is pending successful evaluation.
• Updated the active models table: standardized context window formatting and refreshed TPM figures across the production catalog.
• Added a License type column to Active Models, with links to the license terms for each active model.
• Added a GPU Infrastructure page to the Phoeniqs Cloud platform architecture, covering NVIDIA H100 GPU scheduling, container vs. VM passthrough deployment models, in-VM driver responsibility, default network policy, and tenant monitoring. See the GPU Infrastructure guide.
• Added Model Service Integrations, step-by-step guides for connecting Phoeniqs AI Model Service to existing tools via the OpenAI-compatible API, starting with OpenCode and Cursor. See Model Service Integrations.