Infrastructure / Deployment

Google Cloud Run

Runs any OCI container image as a fully managed Google Cloud service, autoscaling from zero to thousands of instances and billing only while a request is in flight. Google exposes individual Cloud Run instances for long-running background agents, and a built-in `sandbox` tool lets an agent spin up an ephemeral, isolated environment to execute generated code. Container isolation is backed by gVisor.

Google Cloud Run, a fully managed serverless container platform. Scored as the managed deployment counterpart in this category's recipe. Verified 2026-08-13 via cloud.google.com/run, the Cloud Run at Next '26 announcement and the gVisor repository.

Openness

1 high confidence

1.0

license: Proprietary
service: proprietary(Google SaaS)
source: closed
isolation-runtime: gVisor(open source, github.com/google/gvisor) but the Cloud Run product/control-plane is proprietary

Proprietary managed serverless platform; only the underlying gVisor sandbox runtime is OSS, not Cloud Run itself.

https://cloud.google.com/run recorded 2026-08-13
Cloud Run described as fully-managed Google serverless container product (proprietary SaaS)
https://github.com/google/gvisor recorded 2026-08-13
gVisor (Cloud Run's isolation runtime) is the open source component, distinct from the closed product

Adoption

5 high confidence

5.0

Google’s own Next '26 announcement states that external active developers and apps on Cloud Run doubled year over year, and that Cloud Run added more new customers and apps in 2025 than in its first six years combined; a Replit VP is quoted there saying Replit hosts more than 1M live projects with Cloud Run as the primary target. Hyperscaler serverless reach is well into the >10M band. Cloud Run publishes no countable artifact - no package or registry entry - so the band rests on those vendor claims rather than on a recomputable count.

https://cloud.google.com/blog/products/serverless/whats-new-for-cloud-run-at-next26 recorded 2026-08-13
active developers/apps doubled YoY; more new customers in 2025 than first 6 years combined; Replit >1M live projects on Cloud Run

Capability

4 medium confidence

4.0

Very broad runtime/language coverage and hyperscaler scale; the built-in agent `sandbox` tool (announced Next '26, still 'coming soon') adds first-class AI-code-execution via a `sandbox do --` command, alongside Cloud Run instances in preview for long-running background agents. gVisor isolation is process/syscall-level rather than full microVM (Sprites/Northflank), so not the top isolation tier. Strong but not the single frontier-definer. Independent ComputeSDK TTI benchmarks corroborate the sub-second cold-start, placing Cloud Run at a 0.65s median time-to-interactive, 13th by composite of 24 sandbox providers - still sub-second and still mid-field.

https://cloud.google.com/blog/products/serverless/whats-new-for-cloud-run-at-next26 recorded 2026-08-13
built-in `sandbox` tool for agents to run AI-generated code in ephemeral isolated envs; container model, autoscale 0->thousands
https://cloud.google.com/run recorded 2026-08-13
OCI container deploy, scale-to-zero, broad runtime support
https://www.computesdk.com/benchmarks/sandboxes/ recorded 2026-08-13
ComputeSDK TTI leaderboard (independent, 2026-08-07 run, 24 providers): Cloud Run median time-to-interactive 0.65s, P95 1.05s, P99 1.15s, 100% success; 13th by composite score

Verified 2026-08-13