Integration of Red Hat AI Inference Server, validated models, and Llama Stack with Model Context Protocol helps build higher-performance, more consistent AI applications and agents.
Credit: Michael Vi/Shutterstock
At Red Hat Summit 2025, open source solutions company Red Hat announced major updates across its enterprise AI portfolio, including Red Hat AI Inference Server, Red Hat AI third-party verification model, and Llama Stack integration with the Model Context Protocol (MCP) API. Through this, the plan is to continuously expand the options available to companies when deploying generative AI products in a hybrid cloud environment.
According to Forrester, open source software is a key element in accelerating enterprise AI efforts. As the AI environment grows increasingly complex and dynamic, Red Hat AI Inference Server and third-party verification models provide efficient model inference and a collection of verification AI models optimized for the performance of the Red Hat AI Platform. Additionally, Red Hat integrates new APIs for generative AI agent development, including Llama Stack and MCP, to help reduce deployment complexity and accelerate AI initiatives with greater control and efficiency.
The Red Hat AI portfolio includes the new Red Hat AI Inference Server. This server is designed to deliver faster, more consistent, and more cost-effective inference at scale across hybrid cloud environments. This core functionality is integrated into the latest releases of Red Hat OpenShift AI and Red Hat Enterprise Linux AI (RHEL AI) and is also available as a standalone solution to increase deployment flexibility and performance for intelligent applications.
Red Hat AI third-party validation models available through Hugging Face help companies easily find the right model for their specific needs. Red Hat AI provides a collection of validated models and deployment guides to increase customer confidence in model performance and result reproducibility. Some models optimized with Red Hat utilize model compression technology to reduce size and increase inference speed to minimize resource consumption and operating costs.
Red Hat AI integrates the Llama Stack, first developed by Meta, with Antropic’s MCP to provide users with a standardized API for building and deploying AI applications and agents. Llama Stack, now available in developer preview on Red Hat AI, provides a unified API to access vLLM inference, Retrieval-Augmented Generation (RAG), model evaluation, guardrails, and agent functions across all generative AI models. MCP supports integration with external tools in agent workflows by providing standard interfaces to connect APIs, plugins, and data sources.
“Enterprises are moving beyond the initial AI exploration phase and focusing on practical deployments. The key to their continued success will depend on their ability to flexibly adapt their AI strategies to different environments and needs. The future of AI requires not just powerful models, but models that can be deployed proactively and cost-effectively. This flexibility is essential for companies looking to scale their AI initiatives and realize business value,” said Michelle Rosen, research manager at IDC.
If it becomes a problem, I will delete it.
