{"id":54281,"date":"2026-04-23T22:00:00","date_gmt":"2026-04-23T15:00:00","guid":{"rendered":"https:\/\/thaipropertynews.com\/feeds\/?p=54281"},"modified":"2026-04-23T22:00:00","modified_gmt":"2026-04-23T15:00:00","slug":"tencent-cloud-cube-sandbox-goes-fully-open-source-with-five-major-breakthroughs-enabling-large-scale-agent-deployment","status":"publish","type":"post","link":"https:\/\/thaipropertynews.com\/feeds\/?p=54281","title":{"rendered":"Tencent Cloud Cube Sandbox Goes Fully Open-Source, with Five Major Breakthroughs Enabling Large-Scale Agent Deployment"},"content":{"rendered":"<p><i>Tencent Cloud&#8217;s Cube Sandbox goes fully open source with five technical breakthroughs, providing a production-grade foundation for AI Agent deployment at industrial scale.<\/i>\u00a0<\/p>\n<p><span class=\"legendSpanClass\">SHENZHEN, China<\/span>, <span class=\"legendSpanClass\">April 23, 2026<\/span> \/PRNewswire\/ &#8212;\u00a0As AI Agent applications evolve rapidly, building an optimal underlying architecture has become one of the industry&#8217;s most pressing challenges. On April 21, 2026,\u00a0Tencent Cloud officially introduced its open-sourced Cube Sandbox under Apache 2.0\u2014not SDK-only, but the entire production-grade sandbox-as-a-service stack, battle-tested at scale and immediately deployable.<\/p>\n<div class=\"PRN_ImbeddedAssetReference\">\n<p><a href=\"https:\/\/mma.prnasia.com\/media2\/2964133\/image1.html\" target=\"_blank\" rel=\"nofollow\"><img decoding=\"async\" src=\"https:\/\/mma.prnasia.com\/media2\/2964133\/image1.jpg?p=medium600\" title=\"\" alt=\"\" \/><\/a><br \/><span><\/span><\/p>\n<\/div>\n<p><b>A Foundational Layer for the Agent Era<\/b>\u00a0<\/p>\n<p>Cube is the industry&#8217;s only open-source Agent sandbox combining hardware-level isolation with sub-60ms cold start, natively supporting the OpenAI Python SDK and E2B SDK. Developers can simply redirect the runtime and migrate seamlessly. No code changes required.<\/p>\n<p>This gives developers and enterprises a secure, high-performance, low-cost foundation\u2014bringing Agents from the lab into mass production.\u00a0<\/p>\n<p><b>Performance, Security, and Stability at the Limit<\/b><\/p>\n<p>Built at the hardware virtualization layer, Cube Sandbox delivers an extreme combination:<\/p>\n<ul type=\"disc\">\n<li><b>Performance<\/b>: Cold start as low as <b>60ms<\/b> in real-world scenarios\u2014one-third of the industry average (150ms). Minute-level scheduling of tens of thousands of sandboxes, with platform-level burst scheduling exceeding <b>100K instances<\/b>.<\/li>\n<li><b>Security<\/b>: A <b>triple-layer defense architecture<\/b> with millisecond-level event snapshots and state rollback, providing a critical &#8220;undo&#8221; mechanism for unpredictable Agent behavior. This feature will be launched and open-sourced once fully completed.<\/li>\n<li><b>Stability<\/b>: Validated at <b>ultra-large-scale production<\/b>, with every headline performance number measured in live production environments.<\/li>\n<\/ul>\n<p><b>Five Breakthroughs: A Hardware-Level Security Cockpit for Agents<\/b><\/p>\n<p>Built on MicroVM architecture, Cube addresses autonomous Agent security risks: malicious code execution, data exfiltration, resource abuse, and kernel escape.<\/p>\n<div class=\"PRN_ImbeddedAssetReference\">\n<p><a href=\"https:\/\/mma.prnasia.com\/media2\/2964134\/image2.html\" target=\"_blank\" rel=\"nofollow\"><img decoding=\"async\" src=\"https:\/\/mma.prnasia.com\/media2\/2964134\/image2.jpg?p=medium600\" title=\"Figure: Tencent Cloud Cube Sandbox layered architecture\" alt=\"Figure: Tencent Cloud Cube Sandbox layered architecture\" \/><\/a><br \/><span>Figure: Tencent Cloud Cube Sandbox layered architecture<\/span><\/p>\n<\/div>\n<p>At the technical core, Cube Sandbox delivers five breakthroughs:<\/p>\n<ul type=\"disc\">\n<li><b>Hardware-Level Isolation<\/b><\/li>\n<\/ul>\n<p>Every sandbox runs a dedicated Guest OS kernel via KVM hardware virtualization\u2014no shared kernel. &#8220;A breach in one sandbox leaves the rest untouched.&#8221;<\/p>\n<ul type=\"disc\">\n<li><b>Sub-60ms Cold Start<\/b><\/li>\n<\/ul>\n<p>Via resource pool pre-provisioning, snapshot cloning, EPT Lazy Load, and lock optimization, cold start is &lt;60ms (avg. 67ms, P95 = 90ms at 50 concurrent)\u2014significantly outperforming VMs, containers and peer MicroVM solutions.<\/p>\n<ul type=\"disc\">\n<li><b>Extreme Lightweight<\/b><\/li>\n<\/ul>\n<p>Per-instance memory overhead &lt;5MB. CoW sharing, Rust-based trimming, and reflink disk sharing enable 2,000+ sandboxes on a single 96-vCPU host, with 90%+ storage savings versus traditional approaches.<\/p>\n<ul type=\"disc\">\n<li><b>Massive Concurrent Scheduling<\/b><\/li>\n<\/ul>\n<p>Distributed scheduling and bin-packing deliver platform-level burst scheduling of 100K+ instances, with P99 latency below 200ms under 100 concurrent launches on a single 96-vCPU host.<\/p>\n<ul type=\"disc\">\n<li><b>Event-Level Snapshot Rollback<\/b><\/li>\n<\/ul>\n<p>Sub-hundred-millisecond snapshots support checkpoint saving, arbitrary state rollback, and rapid forking. This capability will be released and open-sourced once development is completed.<\/p>\n<p><b>Three User Scenarios Across the Full Agent Lifecycle<\/b><\/p>\n<p>Cube Sandbox addresses the full-lifecycle needs of AI Agents\u2014from R&amp;D and training to application development and enterprise-scale production.<\/p>\n<p><b>For Foundation Model Labs:<\/b><\/p>\n<p>Cube tackles extreme concurrency in Agentic RL training. MiniMax runs hundreds of thousands of heterogeneous sandboxes (Linux, Windows, Android) concurrently. Cube&#8217;s image acceleration cuts storage and IO pressure; distributed scheduling delivers 100K+ instances per minute, several times faster than peer solutions.<\/p>\n<p><b>For Agent Developers and Small-to-Medium Businesses:<\/b><\/p>\n<p>No Kubernetes, no vendor lock-in\u2014a one-click script sets up the full environment in minutes. Integration via MCP, API, SDK, or CLI requires zero rewrites to Agent code.<\/p>\n<p><b>For Enterprise Customers:<\/b><\/p>\n<p>Full private deployment keeps data within enterprise boundaries, meeting cybersecurity grading and compliance requirements. Apache 2.0 ensures commercial friendliness, full auditability, and zero dependency on foreign cloud providers.<\/p>\n<p><b>Already Open, Already in Production<\/b><\/p>\n<p>Cube Sandbox has released its full codebase, one-click deployment scripts, documentation, and examples\u2014covering Shell execution, file operations, browser automation, RL training, and more\u2014free to use, modify, and distribute.<\/p>\n<p>At the <b>Tencent Cloud Shanghai City Summit on March 27<\/b>, Dowson Tong, Senior Executive Vice President of Tencent and CEO of the Cloud and Smart Industries Group, first disclosed the Cube open-source plan as a core practice of Tencent Cloud&#8217;s AI Agent Infrastructure strategy.<\/p>\n<p>Tencent Cloud will pair Cube with TACO AI Acceleration Engine and FlexKV cache system, building a full-stack &#8220;Secure Sandbox + Inference Acceleration + Cache Optimization&#8221; infrastructure\u2014giving every Agent a secure, efficient, low-cost runtime and unlocking the productivity potential of the large-model era.\u00a0<\/p>\n<div class=\"PRN_ImbeddedAssetReference\">  <\/div>","protected":false},"excerpt":{"rendered":"<p><!-- wp:html --><\/p>\n<p><i>Tencent Cloud&#8217;s Cube Sandbox goes fully open source with five technical breakthroughs, providing a production-grade foundation for AI Agent deployment at industrial scale.<\/i>\u00a0<\/p>\n<p><span class=\"legendSpanClass\">SHENZHEN, China<\/span>, <span class=\"legendSpanClass\">April 23, 2026<\/span> \/PRNewswire\/ &#8212;\u00a0As AI Agent applications evolve rapidly, building an optimal underlying architecture has become one of the industry&#8217;s most pressing challenges. On April 21, 2026,\u00a0Tencent Cloud officially introduced its open-sourced Cube Sandbox under Apache 2.0\u2014not SDK-only, but the entire production-grade sandbox-as-a-service stack, battle-tested at scale and immediately deployable.<\/p>\n<div class=\"PRN_ImbeddedAssetReference\">\n<p><a href=\"https:\/\/mma.prnasia.com\/media2\/2964133\/image1.html\" target=\"_blank\" rel=\"nofollow\"><img decoding=\"async\" src=\"https:\/\/mma.prnasia.com\/media2\/2964133\/image1.jpg?p=medium600\" title=\"\" alt=\"\" \/><\/a><br \/><span><\/span><\/p>\n<\/div>\n<p><b>A Foundational Layer for the Agent Era<\/b>\u00a0<\/p>\n<p>Cube is the industry&#8217;s only open-source Agent sandbox combining hardware-level isolation with sub-60ms cold start, natively supporting the OpenAI Python SDK and E2B SDK. Developers can simply redirect the runtime and migrate seamlessly. No code changes required.<\/p>\n<p>This gives developers and enterprises a secure, high-performance, low-cost foundation\u2014bringing Agents from the lab into mass production.\u00a0<\/p>\n<p><b>Performance, Security, and Stability at the Limit<\/b><\/p>\n<p>Built at the hardware virtualization layer, Cube Sandbox delivers an extreme combination:<\/p>\n<ul type=\"disc\">\n<li><b>Performance<\/b>: Cold start as low as <b>60ms<\/b> in real-world scenarios\u2014one-third of the industry average (150ms). Minute-level scheduling of tens of thousands of sandboxes, with platform-level burst scheduling exceeding <b>100K instances<\/b>.<\/li>\n<li><b>Security<\/b>: A <b>triple-layer defense architecture<\/b> with millisecond-level event snapshots and state rollback, providing a critical &#8220;undo&#8221; mechanism for unpredictable Agent behavior. This feature will be launched and open-sourced once fully completed.<\/li>\n<li><b>Stability<\/b>: Validated at <b>ultra-large-scale production<\/b>, with every headline performance number measured in live production environments.<\/li>\n<\/ul>\n<p><b>Five Breakthroughs: A Hardware-Level Security Cockpit for Agents<\/b><\/p>\n<p>Built on MicroVM architecture, Cube addresses autonomous Agent security risks: malicious code execution, data exfiltration, resource abuse, and kernel escape.<\/p>\n<div class=\"PRN_ImbeddedAssetReference\">\n<p><a href=\"https:\/\/mma.prnasia.com\/media2\/2964134\/image2.html\" target=\"_blank\" rel=\"nofollow\"><img decoding=\"async\" src=\"https:\/\/mma.prnasia.com\/media2\/2964134\/image2.jpg?p=medium600\" title=\"Figure: Tencent Cloud Cube Sandbox layered architecture\" alt=\"Figure: Tencent Cloud Cube Sandbox layered architecture\" \/><\/a><br \/><span>Figure: Tencent Cloud Cube Sandbox layered architecture<\/span><\/p>\n<\/div>\n<p>At the technical core, Cube Sandbox delivers five breakthroughs:<\/p>\n<ul type=\"disc\">\n<li><b>Hardware-Level Isolation<\/b><\/li>\n<\/ul>\n<p>Every sandbox runs a dedicated Guest OS kernel via KVM hardware virtualization\u2014no shared kernel. &#8220;A breach in one sandbox leaves the rest untouched.&#8221;<\/p>\n<ul type=\"disc\">\n<li><b>Sub-60ms Cold Start<\/b><\/li>\n<\/ul>\n<p>Via resource pool pre-provisioning, snapshot cloning, EPT Lazy Load, and lock optimization, cold start is &lt;60ms (avg. 67ms, P95 = 90ms at 50 concurrent)\u2014significantly outperforming VMs, containers and peer MicroVM solutions.<\/p>\n<ul type=\"disc\">\n<li><b>Extreme Lightweight<\/b><\/li>\n<\/ul>\n<p>Per-instance memory overhead &lt;5MB. CoW sharing, Rust-based trimming, and reflink disk sharing enable 2,000+ sandboxes on a single 96-vCPU host, with 90%+ storage savings versus traditional approaches.<\/p>\n<ul type=\"disc\">\n<li><b>Massive Concurrent Scheduling<\/b><\/li>\n<\/ul>\n<p>Distributed scheduling and bin-packing deliver platform-level burst scheduling of 100K+ instances, with P99 latency below 200ms under 100 concurrent launches on a single 96-vCPU host.<\/p>\n<ul type=\"disc\">\n<li><b>Event-Level Snapshot Rollback<\/b><\/li>\n<\/ul>\n<p>Sub-hundred-millisecond snapshots support checkpoint saving, arbitrary state rollback, and rapid forking. This capability will be released and open-sourced once development is completed.<\/p>\n<p><b>Three User Scenarios Across the Full Agent Lifecycle<\/b><\/p>\n<p>Cube Sandbox addresses the full-lifecycle needs of AI Agents\u2014from R&amp;D and training to application development and enterprise-scale production.<\/p>\n<p><b>For Foundation Model Labs:<\/b><\/p>\n<p>Cube tackles extreme concurrency in Agentic RL training. MiniMax runs hundreds of thousands of heterogeneous sandboxes (Linux, Windows, Android) concurrently. Cube&#8217;s image acceleration cuts storage and IO pressure; distributed scheduling delivers 100K+ instances per minute, several times faster than peer solutions.<\/p>\n<p><b>For Agent Developers and Small-to-Medium Businesses:<\/b><\/p>\n<p>No Kubernetes, no vendor lock-in\u2014a one-click script sets up the full environment in minutes. Integration via MCP, API, SDK, or CLI requires zero rewrites to Agent code.<\/p>\n<p><b>For Enterprise Customers:<\/b><\/p>\n<p>Full private deployment keeps data within enterprise boundaries, meeting cybersecurity grading and compliance requirements. Apache 2.0 ensures commercial friendliness, full auditability, and zero dependency on foreign cloud providers.<\/p>\n<p><b>Already Open, Already in Production<\/b><\/p>\n<p>Cube Sandbox has released its full codebase, one-click deployment scripts, documentation, and examples\u2014covering Shell execution, file operations, browser automation, RL training, and more\u2014free to use, modify, and distribute.<\/p>\n<p>At the <b>Tencent Cloud Shanghai City Summit on March 27<\/b>, Dowson Tong, Senior Executive Vice President of Tencent and CEO of the Cloud and Smart Industries Group, first disclosed the Cube open-source plan as a core practice of Tencent Cloud&#8217;s AI Agent Infrastructure strategy.<\/p>\n<p>Tencent Cloud will pair Cube with TACO AI Acceleration Engine and FlexKV cache system, building a full-stack &#8220;Secure Sandbox + Inference Acceleration + Cache Optimization&#8221; infrastructure\u2014giving every Agent a secure, efficient, low-cost runtime and unlocking the productivity potential of the large-model era.\u00a0<\/p>\n<div class=\"PRN_ImbeddedAssetReference\">  <\/div>\n<p><!-- \/wp:html --><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"rop_custom_images_group":[],"rop_custom_messages_group":[],"rop_publish_now":"initial","rop_publish_now_accounts":[],"rop_publish_now_history":[],"rop_publish_now_status":"pending","footnotes":""},"categories":[5,7],"tags":[],"class_list":["post-54281","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-cision-pr-newswire","category-cision-pr-newswire-en"],"_links":{"self":[{"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=\/wp\/v2\/posts\/54281","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=54281"}],"version-history":[{"count":0,"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=\/wp\/v2\/posts\/54281\/revisions"}],"wp:attachment":[{"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=54281"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=54281"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/thaipropertynews.com\/feeds\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=54281"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}