CNCF Announces General Availability of Dapr Agents v1.0 for Production AI Workloads

The Cloud Native Computing Foundation announced the general availability of Dapr Agents v1.0, a Python framework built on Dapr’s distributed application runtime that helps teams run reliable and secure AI agents in production environments.

The release marks the transition from early experimentation to stable production use and is the result of a yearlong collaboration between NVIDIA, the Dapr open source community and end users building practical AI agent systems. Dapr Agents v1.0 provides durable workflows, state management and secure multi-agent coordination for teams deploying AI agents on Kubernetes and cloud native platforms.

Key capabilities include durable long-running agent workflows with automatic retries and failure recovery, persistent state across more than 30 databases, secure communication and identity using SPIFFE, multi-agent coordination and messaging, and built-in observability and monitoring. The framework also offers flexibility to switch language model providers without code changes.

“The Dapr Agents v1.0 milestone provides the essential cloud native guardrails, like state management and secure communication, that platform teams need to turn AI prototypes into reliable, production-ready systems at scale,” said Chris Aniszczyk, CTO of CNCF.

“Many agent frameworks focus on logic alone,” said Mark Fussell, Dapr maintainer and steering committee member. “Dapr Agents delivers the infrastructure that keeps agents reliable through failures, timeouts and crashes. With v1.0, developers have a foundation they can trust in production.”

At KubeCon + CloudNativeCon Europe, ZEISS Vision Care will present a real-world implementation using Dapr Agents to extract optical parameters from highly variable, unstructured documents.