Red Hat announced new products and partnerships across a wide range of areas, including strengthening its enterprise AI platform, space computing and software-defined vehicles, at its annual Red Hat Summit 2026 conference held on May 11 local time.

First, it highlighted the 'Red Hat AI 3.4' update.

Red Hat AI 3.4 supports large-scale inference and agentic AI deployments. It adds model service capabilities that manage access to AI models through a central gateway, track usage and apply policies.

It also introduced a 'speculative decoding' technique that increases inference speed by up to three times. Agent management and observability were strengthened, and support was added for a Model Context Protocol (MCP) gateway and a catalog.

Joe Fernandez (조 페르난데스), vice president of Red Hat AI, said, "Inference will be the dominant workload in enterprise AI." He said, "AI agents will increase inference demand exponentially. Red Hat is focused on enabling enterprises to connect their own data to existing models rather than building foundation models from scratch."

Red Hat also expanded cooperation with Nvidia. It added support for Nvidia's Blackwell architecture and Vera Rubin platform, and decided to participate in Nvidia's OpenShell project for AI agent sandboxing and secure execution.

Red Hat also announced it will work with Voyager Technologies to deploy Red Hat Enterprise Linux 10.1 to an International Space Station (ISS) space edge micro data centre. The goal is to process data on site and support AI workloads without sending the data to Earth in an environment with limited power and hardware.

Red Hat will work with Japanese automaker Nissan to develop a next-generation software-defined vehicle platform. The companies announced a joint engineering initiative at the event using the Red Hat In-Vehicle Operating System. Through this, Red Hat will provide a standardised Linux-based foundation for Nissan's vehicle computer architecture and support software updates and AI functions across the vehicle lifecycle.