Anthropic Launches Claude Managed Agents for Enterprise AI Workflows
Anthropic debuts Claude Managed Agents, a fully managed cloud service that lets developers build and deploy autonomous AI agents with secure sandboxing, persistent sessions, and built-in tooling — no infrastructure required.
Anthropic has launched Claude Managed Agents, a fully managed cloud service designed to take the infrastructure burden out of building autonomous AI agents. Announced on April 8, 2026, the service provides developers with a complete agent harness — secure sandboxed code execution, authentication, session checkpointing, scoped permissions, and server-sent event streaming — all built on top of Claude models. Access is available in public beta to all Anthropic API accounts via the managed-agents-2026-04-01 header.
The service targets enterprise teams that want to move quickly from prototype to production without building and maintaining the underlying technical layers themselves. Developers can deploy agents through the Claude Console, Claude Code, or the newly introduced Managed Agents CLI. Pricing is set at $0.08 per session hour on top of standard Claude API token costs, a model that aligns the cost of compute with actual agent runtime rather than flat subscription tiers.
High-profile early adopters already integrated into Claude Managed Agents include Notion, Asana, Rakuten, Sentry, and Vibecode. These companies are leveraging the platform to power automated workflows ranging from document processing and project management to customer support and code review. The persistent long-running session support is a key differentiator, enabling agents to handle complex multi-step tasks that unfold over minutes or hours without losing state.
The launch positions Anthropic squarely in the growing market for managed AI agent infrastructure, where enterprises want the power of autonomous AI without the operational complexity. By abstracting away sandboxing, permissions, and session management, Anthropic is betting that the bottleneck for enterprise AI adoption is not model capability but deployment reliability. With Claude Managed Agents, the company is making a strong play to become the preferred backend for production-grade AI agents across industries.
