Gmail Reply Draft Automation with n8n + Groq Llama 3.3

Automatically draft smart Gmail replies with n8n + Groq (Llama 3.3)

This n8n workflow polls your Gmail for new messages, uses Groq’s Llama 3.3 to decide whether a human reply is needed, and then creates a Gmail draft reply only when appropriate—so you stay in control without sending anything automatically.

What this workflow does

• Monitors Gmail every minute for newly received emails using a Gmail Trigger.
• Evaluates each email with Groq Llama 3.3: it sends the email text (up to 3,000 characters) to the Groq Chat Completions API via an HTTP Request.
• Receives a JSON response and parses it to extract:
  • the model’s decision on whether a reply is needed
  • the suggested draft reply text
  • the original subject
• Creates a Gmail draft reply with a subject like “Re: <original subject>” when the message is marked as needing a response.

Nothing is sent automatically—this workflow only drafts replies when the model determines human follow-up is required.

Use cases

• Support inbox triage: draft replies for emails that need an answer, while ignoring non-actionable messages.
• SaaS operations: generate consistent first-draft responses from incoming customer inquiries.
• Inbox efficiency: reduce time spent deciding whether to respond by using an LLM-based “reply needed” decision.

Technical details

• n8n nodes/workflow components: Gmail Trigger, Gmail (Draft) node, HTTP Request, Code, If, Sticky Note.
• Groq integration: send email content to Groq’s Llama 3.3 70B Chat Completions API using your Groq API key (Bearer token).
• Setup requirements: Gmail OAuth2 credential, Groq API key from console.groq.com, optional system prompt customization to match your support tone and rules.

Perfect for n8n users and automation engineers building practical Gmail reply automation with LLM-assisted decisioning.