Google Cloud Run (Compute)
- Free allowance
- Per billing account, monthly: 2 million requests, 180,000 vCPU-seconds and 360,000 GiB-seconds under request-based billing; 240,000 vCPU-seconds and 450,000 GiB-seconds under instance-based billing; plus 1 GiB of outbound transfer from North America
- Verdict
- caveatusable, with a named limit or an unprobed read path
- Cost
- free tier
- Identity
- cardcard required, no phone
- Binding limit
- volume-quota240,000 vCPU-seconds and 450,000 GiB-seconds a month under instance-based billing
- Evidence
- documented8 sources; a first-party page was read and the sentence quoted
- Freshness
- stale on 2026-12-05checked 2026-11-04 (live); stale on 2026-12-05
| Free tier | per billing account, monthly: 2 million requests, 180,000 vCPU-seconds and 360,000 GiB-seconds under request-based billing; 240,000 vCPU-seconds and 450,000 GiB-seconds under instance-based billing; plus 1 GiB of outbound transfer from North America unit seconds; metered on wall-clock |
|---|---|
| Price | none stated |
| Limits |
it was not resolved either |
| Binding | volume-quota an always-on instance spends the monthly grant in about ten days: 450,000 GiB-seconds at 512 MiB is 900,000 seconds, 10.4 days. |
| Identity | card a card at signup; the free tier is a spending-based discount applied after the fact, not a prepaid allowance |
| Egress | 1 GiB of data transfer within North America free per month, then Google Cloud network pricing at the Premium tier free allowance: 1 GiB |
| Per-object cap | a request may run up to 60 minutes; memory and CPU per instance are configurable |
| Retention | no persistent disk; a container is stateless unless it mounts a volume or a bucket |
| Interface | rest, cli an OCI image behind an HTTPS endpoint; the service listens on a port |
| Tooling | gcloud, the Cloud Run API, Docker; any language with an HTTP server |
| Composes with | the only free serverless tier here with 50 vCPU-hours and a listening socket |
| Reliability | a service scales to zero by default and pays a cold start on the next request; a minimum instance removes the cold start and starts the idle meter |
| Notes | Arithmetic from the published free tiers, not a measurement: at 0.25 vCPU and 512 MiB, an always-on instance needs 648,000 vCPU-seconds and 1,296,000 GiB-seconds a month. The memory grant runs out first, at 900,000 seconds, so the free tier holds an always-on 0.25 vCPU service for about 10.4 days a month. What is not verified is whether the 'Idle time (Min instance)' SKU is drawn from the same free tier; the page presents it as a separate rate. |
| Evidence note | documented both free-tier tables and the per-second rates were read on the pricing page |
| Parsed numbers |
|
Evidence 8
| date | source | what it proved |
|---|---|---|
| 2026-10-22 | cloud.google.com/free/docs/free-cloud-features | Cloud Run free tier: 2 million requests, 360,000 GB-seconds of memory, 180,000 vCPU-seconds, 1 GB of outbound transfer a month |
| 2026-10-24 | cloud.google.com/run/pricing | quoted: 'Free tier (based on us-central1 pricing): CPU - First 240,000 vCPU-seconds free per month; RAM - First 450,000 GiB-seconds free per month' under instance-based billing |
| 2026-10-24 | cloud.google.com/run/pricing | quoted: 'CPU - First 180,000 vCPU-seconds free per month; RAM - First 360,000 GiB-seconds free per month; Requests - 2 million requests free per month' under request-based billing |
| 2026-10-24 | cloud.google.com/run/pricing | quoted: 'The free tier usage is aggregated across projects by billing account and resets every month' |
| 2026-10-24 | cloud.google.com/run/pricing | quoted rate table: 'CPU (per vCPU-second) ... $0.000018', 'Memory (per GiB-second) ... $0.000002', 'Idle time (Min instance1) $0.0000025' |
| 2026-11-04 | docs.cloud.google.com/sdk/gcloud/reference/run/deploy | a gcloud CLI command deploys Cloud Run services |
| 2026-11-04 | docs.cloud.google.com/run/docs/reference/rest | Cloud Run has a documented REST API |
| 2026-11-04 | cloud.google.com/run/pricing | outbound data transfer has a 1 GiB monthly free allowance in North America, then network pricing |
More fields 15
| keep rate | {"value": 0.0,"basis": "a scaled-to-zero service does not bill","source": "https://cloud.google.com/free/docs/free-cloud-features"} |
|---|---|
| billing model | per-100ms |
| shape | {"vcpu": 0.25,"ram_gib": 0.5,"disk_gib": null,"gpu_class": null,"gpu_count": null} |
| session cap | {"wall_clock_min": 60,"idle_reap_s": null,"restart_required": false,"note": "a request may run 60 minutes; the service scales to zero when idle unless a minimum instance is set"} |
| persistent storage | {"included_gib": 0,"price_gib_month": "Cloud Storage is separate","durability": "none in the runtime","survives_stop": false,"survives_destroy": false,"max_retention": "none"} |
| network out | {"allow_outbound": true,"allow_inbound_listen": true,"ports": "the container port on a generated HTTPS hostname","static_ip": false,"ipv6": false,"per_session_bytes": null} |
| granularity | {"min_billed_s": 0.1,"rounding_s": 0.1} |
| free tier kind | monthly-spending-discount |
| open question | whether the grant applies to the Services meter or the Instances meter |
| gap | the grant's meter (Services or Instances) is not resolved; the claim page's Instances section and footnotes would close it |
| burst x | 1.0 |
| term | monthly |
| m10h usd | 0.0 |
| m24h usd | 0.0 |
| price arithmetic | free within the monthly grant; 512 MiB at 0.25 vCPU consumes the 450,000 GiB-second grant in 900,000 s = 10.4 days of 24/7 |