Two major players in the open transformation of artificial intelligence have just announced an extension of their collaboration to advance AI. The progress in the partnership between Red Hat and Google Cloud focuses on expanding the range of enterprise applications for AI, combining the open-source technologies of the open-source leader with Google’s specialized infrastructure and its family of models, Gemma.
Together, the companies will advance use cases for scaling AI through:
- Launch of the open-source project llm-d, with Google as a founding collaborator;
- Support for vLLM on Google Cloud TPUs and GPU-powered virtual machines (VMs) to enhance AI inference;
- Day-zero support for vLLM with Gemma 3 model distributions;
- Availability of Red Hat AI Inference Server on Google Cloud;
- Development of agentic AI with Red Hat as a collaborator in Google’s Agent2Agent (A2A) protocol.
Strengthening AI inference with vLLM
Demonstrating its commitment to readiness from day zero, Red Hat is now one of the first testers of Google’s family of open models, Gemma, starting with Gemma 3, with immediate support for vLLM. vLLM is an open-source inference server that accelerates the execution of generative AI applications. As a key commercial collaborator of vLLM, Red Hat is making this platform more efficient and responsive for gen AI applications.
Additionally, Google Cloud TPUs, high-performance AI accelerators integrated into Google’s AI portfolio, are now fully compatible with vLLM. This integration enables developers to maximize resources while achieving the performance and efficiency essential for fast and accurate inference.
Recognizing the transition from AI research to real-world deployment, many organizations face the complexities of a diverse AI ecosystem and the need to adopt more distributed computing strategies. To meet this demand, Red Hat launched the open-source project llm-d, with Google as a founding collaborator. Leveraging the momentum of the vLLM community, this initiative aims to usher in a new era for gen AI inference. The goal is to enable greater scalability across heterogeneous resources, optimize costs, and improve workload efficiency—all while fostering continuous innovation.
Driving enterprise AI with community-based innovation
Bringing the latest advancements from the open-source community to the enterprise environment, the Red Hat AI Inference Server is now available on Google Cloud. Just like Red Hat’s enterprise distribution of vLLM, the AI Inference Server helps businesses optimize model inference across their hybrid cloud environment. Using Google Cloud’s reliable infrastructure, organizations can deploy production-ready generative AI models that are both highly responsive and cost-effective at scale.
Highlighting their joint commitment to open AI, Red Hat has also begun contributing to Google’s Agent2Agent (A2A) protocol —an application-level protocol that facilitates communication between agents and end-users across diverse platforms and clouds. By actively participating in the A2A ecosystem, Red Hat aims to accelerate innovation and ensure AI workflows remain dynamic and effective with the power of Agentic AI.
Red Hat Summit
Watch the Red Hat Summit keynotes to hear the latest updates from Red Hat executives, customers, and partners:
- Modern infrastructure aligned with enterprise AI—Tuesday, May 20, 8 AM–10 AM EDT (YouTube)
- Hybrid cloud evolves to drive enterprise innovation—Wednesday, May 21, 8 AM–9:30 AM EDT (YouTube)
