Gloo Gateway 1.19 accelerates context-rich, real-time AI apps with Gateway API

We’re excited to announce Gloo Gateway 1.19 release is available now. This release introduces advanced AI capabilities for users deploying production-grade AI applications, including support for streaming prompt guardrails and tracing ensuring secure, real-time responsiveness between their applications and large language models (LLMs), alongside function calling and tool support helping developers connect their services and applications data better.

This release also continues Gloo Gateway’s close alignment with the Kubernetes Gateway API, with support now available for Gateway API v1.2.1, ensuring up-to-date interoperability with Kubernetes standards across workloads.

Advancing Real-Time AI Application User Experiences 

In Gloo Gateway 1.18, we announced the general availability of Gloo AI Gateway, which included support for upstream large language model (LLM) providers and support for advanced guardrails. In this latest 1.19 release, we’ve expanded these capabilities by adding support for streaming prompt guardrail, tracing, and function and tool calling.

As demand for AI applications grows, users increasingly expect faster, real-time interactions. Streaming mode helps meet these expectations by reducing output latency and delivering results as they are generated, improving the perceived user experience and responsiveness. To ensure these real-time outputs remain secure and performant, application developers can utilize the new streaming prompt guardrails and tracing support in Gloo AI Gateway to provide the same consistent level of control and observability across non-streaming workload requests.

Simultaneously, applications are now expected to deliver richer outputs and more contextual responses beyond static LLM outputs. With Gloo AI Gateway support for function and tool calling, developers can integrate their application’s LLMs with external APIs and services to help unlock and improve real-time data retrieval, customize actions, and enrich their LLM response context across streaming and non-streaming modes. 

You can learn more about Gloo AI Gateway features in the documentation for Gloo Gateway.

Expanding Kubernetes Gateway API Support

We’ve remained committed to closely aligning Gloo Gateway with the latest Kubernetes Gateway API versions, giving users access to the latest features to maintain Kubernetes standards. In Gloo Gateway 1.19, we’ve added support for Kubernetes Gateway API version 1.21, introducing key features including:

  • TLSRoutes – Enables TLS passthrough on TLS listeners via TLSRoute.
  • Multiple TargetRefs – Allows attaching multiple TargetRefs across both Listeners and HTTPRoutes.
  • Backend Protocol Selection – Automatically detects the protocol for gateway-to-backend communication using Kubernetes Standard Application Protocols.
  • ListenerSets – Supports the Gateway API 1.3 XListenerSet feature, decoupling listener config from the gateway and substantially increasing the number of listens per gateway.

Gloo Gateway 1.19 is available now!

Gloo Gateway users can start taking advantage of the new features of Gloo Gateway 1.19 now. Check out the latest release notes and documentation to get started.

Cloud connectivity done right