Unlocking agentic AI: Secure Sandboxes are officially GA @cloudflare
Unlocking agentic AI: Secure Sandboxes are officially GA  @cloudflare
Uploaded April 2026 | Updated November 2026, 1 day ago
Join Mike Nomitch, Product Manager on the Containers team, and Kate Reznykova, Engineering Manager on the Agents and Sandboxes team, as they explain why AI agents need full computer environments to run complex code.

Tune in to learn about these three major updates:

* State Preservation: Save full disk and memory states with R2 snapshots, allowing agents to instantly resume tasks without reinstalling packages.

* Secure Egress: Programmatic proxies and outbound workers dynamically inject credentials, ensuring secrets are never exposed to the agent.

* Limits & Pricing: Concurrently run up to 15,000 lite instances and only pay for active CPU cycles.

Read the blog posts:
blog.cloudflare.com/sandbox-ga

Visit cloudflare.com/agents-week for every announcement and CFTV episode — check back all week for more!
Unlocking agentic AI: Secure Sandboxes are officially GACloudflare Immerse Istanbul 2026Cloudflare Mesh: Secure private networkingDemo: Prevent data exposure in AIDemo: Discover workforce use of shadow AITop Hackers vs AI Agents — Chema Alonso Explains What ChangedCloudflare Agents Week Preview: What to ExpectCloudflare Radar Reels Ep 10 – Tracking Traffic and Connection QualityImmerse Amsterdam After MovieHow This Bank CISO Can Redeploy His Infrastructure in MinutesCloudflare’s New CMO Jeff Samuels: Why Brand Wins, Marketers as Context EngineersAgile SASE | Fast path to safe AI adoption
Cloudflare |

Unlocking agentic AI: Secure Sandboxes are officially GA

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER