# Baseten: Major outage in GCP with partial impact on several customers

- Impact (vendor’s label): none
- Status: resolved
- Started: 2025-06-12 18:17 UTC · Resolved: 2025-06-12 18:33 UTC · Duration: 16 min
- Updates: 1
- [Vendor status page: https://status.baseten.co/incidents/p4s4zsjhg198](https://status.baseten.co/incidents/p4s4zsjhg198)
- [Company: Baseten](https://rewindex.com/company/baseten.md)

## Part of an outage event

- [Google Cloud outage, Jun 12, 2025: 101 vendors, 107 incidents](https://rewindex.com/events/2025-06-12-google-cloud)
- This incident blames Google Cloud as the cause. (hop 1, posted +22 min vs the first vendor naming Google Cloud)
- Evidence: “…Major outage in GCP with partial impact on several customers All inference workloads are operational due to our multi-cloud capacity management. Information here: https://www.baseten.co/blog/how-baset…”

```mermaid
flowchart LR
  root["Google Cloud<br/>own incident not archived"]:::inferred
  h1["Within an hour<br/>29 vendors, 28 blamed Google Cloud<br/>Pinecone, Northflank, imgix +26"]:::blames
  root --> h1
  h4["1 to 4 hours later<br/>12 vendors, 11 blamed Google Cloud<br/>Smartsheet, Daily, Confluent +9"]:::blames
  root --> h4
  later["Later<br/>1 vendor, 1 blamed Google Cloud<br/>Zilliz"]:::blames
  root --> later
  p0["Firebase<br/>own incident not archived"]:::inferred
  root -.-> p0
  k0["Codeium, Windsurf<br/>+13 min"]:::hop2
  p0 -->|"blamed Firebase"| k0
  p1["Anthropic<br/>posted +28 min, did not name Google Cloud"]:::timing
  root -.-> p1
  k1["GitHub, GitHub Copilot<br/>+24 min"]:::hop2
  p1 -->|"blamed Anthropic"| k1
  p2["Cloudflare<br/>own incident not archived"]:::inferred
  root -.-> p2
  k2["Doppler, Jupyter (Anaconda), Socket +4<br/>+35 min to +3 h 41 min"]:::hop2
  p2 -->|"blamed Cloudflare"| k2
  me["Baseten (this incident)<br/>posted +22 min"]:::me
  h1 --- me
```

_Detected automatically from incidents that started together and named the same provider; a link means one incident named another vendor, not a confirmed cause._

## Timeline of updates

- **resolved** 2025-06-12 18:33 UTC: All inference workloads are operational due to our multi-cloud capacity management. Information here: https://www.baseten.co/blog/how-baseten-multi-cloud-capacity-management-mcm-powers-cloud-self-hosted-and-hybr/

_From the vendor’s published history: only the final update text is available for incidents before Rewindex began polling this page._


---
Canonical: https://rewindex.com/company/baseten/incidents/p4s4zsjhg198 · Markdown: https://rewindex.com/company/baseten/incidents/p4s4zsjhg198.md · JSON: https://rewindex.com/api/v1/companies/baseten/incidents/p4s4zsjhg198 · Generated 2026-10-20T00:07:38.358Z