Frequently Asked Question
1: Why is my AI Vision/OCR workflow failing or hanging mid-execution? (Token & Credit Limits)
Last Updated about a month ago
Problem: The automation workflow suddenly stops processing documents, leaves files unhandled, or outputs API timeout/quota errors.
Likely Cause: Your independent Google Cloud Platform (GCP) project has run out of billing credits, or the files uploaded have breached your model's single-request token limits.Immediate Resolution Steps:
- Check Your Private Google Cloud Console: Because you own and manage your own cloud tokens, our support team cannot view or manage your billing status. Log directly into your Google Cloud Console (
://google.com) and navigate to the Billing section. - Verify Free Credit Status: If your project was running on the standard $300.00 USD free tier credits, verify that these credits have not expired or been entirely consumed by heavy OCR data extraction.
- Resolve Billing Holds: If your credits are at zero, Google will immediately block your local engine from communicating with the Gemini model. You must add or update a valid corporate payment method inside your GCP console to restore API functionality.
- Verify Document Token Volume: Large, multi-page PDFs or high-resolution image batches require millions of input tokens. If your billing is active but specific files fail, split the files into smaller batches to prevent breaching the model’s single-request token volume limits.