503 error on ai pro acount

Google Antigravity
1

BugBug HTTP 503 MODEL_CAPACITY_EXHAUSTED on specific account only — all other AI Pro accounts work fine

Problem Description

I’m consistently getting HTTP 503 MODEL_CAPACITY_EXHAUSTED errors on one specific Google AI Pro account. Other accounts on the same AI Pro plan, on the same machine, work perfectly with the same models. This suggests an account-level provisioning or routing issue rather than actual capacity exhaustion.

Error Response

json

{

“error”: {

“code”: 503,

“details”: [

{

“@type”: “type.googleapis.com/google.rpc.ErrorInfo”,

“domain”: “cloudcode-pa.googleapis.com”,

“metadata”: {

“error_number”: “2010”,

“model”: “claude-opus-4-6-thinking”

},

“reason”: “MODEL_CAPACITY_EXHAUSTED”

}

],

“message”: “No capacity available for model claude-opus-4-6-thinking on the server”,

“status”: “UNAVAILABLE”

}

}

Key Details

  • TraceID: 0x7233018b58a.....
  • Trajectory ID: c8408d15-10ea-4a32-9959-59f367......
  • Endpoint: cloudcode-pa.googleapis.com (correct production endpoint)
  • Model: claude-opus-4-6-thinking (happens with other models too)
  • Plan: Google AI Pro on all accounts
  • OS: Windows 11

Why this is a bug, not a capacity issue

  • Multiple accounts, all on AI Pro, same machine
  • Only one specific account consistently returns 503
  • Other accounts work fine with the exact same models at the exact same time
  • Plenty of quota remaining on the affected account

This appears to be an account-level routing or provisioning issue on the backend. Requesting investigation into why this specific account is being routed to exhausted capacity pools while other identical-tier accounts are not.