Telegram Photo Validation with OpenAI Vision + Google Sheets Log

Validate Telegram Photos with OpenAI Vision—Log Every Result to Google Sheets

This n8n workflow automatically reviews photos sent to your Telegram bot using OpenAI Vision, returns a clear pass/fail verdict with confidence and reasons, and records each submission in a Google Sheets audit log.

What this workflow does

• Listens on Telegram for incoming messages sent to your bot (Telegram Trigger). It uses the attached image (largest size).
• Ensures a photo is included: if the message has no image, the bot replies asking the user to send an image.
• Runs OpenAI Vision validation: the workflow sends the photo plus your configured validation criteria to an OpenAI vision-capable chat model and requests a structured JSON verdict.
• Parses the model output into fields such as validity, confidence, reasons, and a neutral photo description (including auto-fixing if the JSON needs correction).
• Applies your rules: it verifies the photo is marked valid and that the confidence score meets your configured threshold.
• Responds in Telegram: the user gets a success/failure reply with the description/reasons and a confidence percentage.
• Logs to Google Sheets: every validation outcome is appended to a specified spreadsheet and sheet for tracking and auditability.

Use cases

• Moderate user-submitted images in a Telegram-based intake process (e.g., proof of requirements).
• Quality-check photos against specific criteria before your team reviews them.
• Create a consistent evidence trail by storing outcomes and confidence in Google Sheets.

Technical details

• Telegram Trigger + Telegram reply nodes to interact with users.
• OpenAI Vision (vision-capable chat model such as gpt-5-mini) to generate a JSON verdict.
• Google Sheets OAuth2 integration to append validation results (audit log).
• Control flow using if, set, and supporting nodes (including a Sticky Note for clarity).