Title: Persistent 429 RESOURCE_EXHAUSTED for 5+ days despite 5-hour quota timer resetting
Hi team,
I’ve been getting “Agent execution terminated due to error” on every single prompt (including trivial ones like “hi”) since last Sunday — that’s over 5 days now. The Models settings page shows all quota bars near-empty with “Refreshes in 4 hours, 59 minutes,” but even after that timer completes and resets, the error persists immediately on the next prompt.
Debug info from the error card:
Trajectory ID: cc87864f-d41c-40de-9c36-5*****
X-Cloudaicompanion-Trace-Id: 4cd9b04b8c1*****
Full error response:
{
“error”: {
“code”: 429,
“message”: “Resource has been exhausted (e.g. check quota).”,
“status”: “RESOURCE_EXHAUSTED”
}
}
Context: I did a burst of heavy development/usage right before this started on Sunday. Since then, every model (including Claude Sonnet 4.6, Claude Opus 4.6, and GPT-OSS 120B) returns the same 429 immediately, even in a fresh conversation.
The 5-hour rolling quota timer keeps completing and resetting, but the error never clears, which makes me think there’s a separate, longer-term cap (weekly or similar) sitting underneath the advertised 5-hour window that isn’t shown anywhere in the UI.
Could someone from the team take a look at my account’s quota state directly? Happy to provide any additional diagnostics needed.
Thanks.