[Inference] 5xx errors on some `llama-text-embed-v2` requests
DegradedIncident is resolved.
Is Pinecone down?
Pinecone is up and running. Pinecone runs a hosted vector database for AI applications.
Last 90 days
99.58%
Estimated uptime
Estimated uptime
90 days agoToday
Worked out from the incidents Pinecone reported, so it is an estimate and not a measurement.
Pinecone reliability, last 90 days
7Incidents reported
0Full outages
58 minTypical duration
11 h 7 minLongest incident
Last incident: “[Inference] 5xx errors on some `llama-text-embed-v2` requests”, 8 days ago.
Components
us-west1-gcpOperational
ConsoleOperational
AWS us-west-2Operational
us-east1-gcpOperational
AWS us-east-1Operational
us-central1-gcpOperational
AWS eu-west-1Operational
us-west4-gcpOperational
Azure eastus2Operational
us-east4-gcpOperational
GCP us-central1Operational
us-west1-gcp-freeOperational
GCP europe-west4Operational
us-west4-gcp-freeOperational
gcp-starterOperational
eu-west1-gcpOperational
eu-west4-gcpOperational
asia-southeast1-gcpOperational
asia-northeast1-gcpOperational
asia-southeast1-gcp-freeOperational
Index ManagementOperational
northamerica-northeast1-gcpOperational
InferenceOperational
eastus-azureOperational
AssistantOperational
us-east-1-awsOperational
Recent incidents
[SERVERLESS][GCP][us-central1] Newly created indexes failing to initialize
DegradedThis incident has been resolved.
[Pods][AWS][us-east-1] Write failures and high latency
DegradedThe third party dependency recovered, this incident has resolved.
[Serverless][AWS][eu-west-1] 5xx on read requests
DegradedThis incident has been resolved.
[Serverless][AWS][us-west-2] 5xx errors on the readpath
DegradedThis incident has been resolved.
Questions
Is Pinecone down right now?
Pinecone’s official status page says: All Systems Operational. This page checks it again when you load it, at most every 5 minutes.
When was Pinecone last down?
The most recent incident Pinecone reported was “[Inference] 5xx errors on some `llama-text-embed-v2` requests”, 8 days ago (2026-10-14 14:48 UTC). None of its incidents in the last 90 days were a full outage.
Which parts of Pinecone are tracked here?
Pinecone reports 26 components on its status page, including us-west1-gcp, Console, AWS us-west-2, us-east1-gcp, AWS us-east-1. Each is checked again whenever this page refreshes.
What should I do if Pinecone is down?
Retry with exponential backoff, fall back to another model or provider if your product allows it, and avoid hammering the API while it recovers.
Pinecone down — does it hit every index?
No — Pinecone lists indexes per cloud and region (AWS us-east-1, GCP us-central1, Azure…), and serverless vs pod indexes are separate rows. Match the row to where your index lives.
Free-tier indexes down but paid ones fine?
Possible — the status page lists free-tier environments separately (rows like 'us-west1-gcp-free'). Free-tier incidents don't imply paid regions are affected.
How do I tell if Pinecone is down for everyone or only for me?
If the status above is operational but Pinecone still fails for you, the cause is probably on your side: try another network, clear the cache, or check your own connection. If it shows a degradation or an outage, others are affected too.
How can I get alerted when my own site goes down?
OpenUptime checks your site every minute and sends an alert to Slack, Telegram, email or a webhook the moment it stops responding. It is open source and free to start.