AI Weekly Newsletter Automation in n8n: Sheets + Tavily + OpenRouter

Automate your weekly newsletter end-to-end in n8n—research with Tavily, write with OpenRouter (OpenAI), and email it from Google Sheets

This n8n workflow automatically finds (or selects) a newsletter topic, gathers three recent news results via Tavily, drafts a complete three-section newsletter with OpenRouter (OpenAI) models, formats it as HTML, and sends it to your subscriber list through Gmail—every week.

What this workflow does

• Runs on a weekly schedule using a schedule trigger.
• Reads Google Sheets for rows where Action = "Generate" and uses the provided Newsletter Topic when present.
• If no manual topic is provided, it searches Reddit via Tavily and uses an OpenRouter (OpenAI) model to pick a single newsletter topic, then writes that topic back to Google Sheets.
• Generates a unique hex ID, updates the queued row in Google Sheets with the ID/source, and performs Tavily research to fetch three recent news results.
• Uses OpenRouter (OpenAI) to plan an issue title and exactly three section topics, then runs Tavily searches per topic and writes each section with citations.
• Merges the sections into an email-ready subject and body, saves the final title/content, and marks the row as DONE ✅.
• Converts the newsletter to styled HTML, reads recipients from a separate Google Sheets leads file, and sends via Gmail with a short wait between messages.

Use cases

• Weekly SaaS updates: generate consistent, cited newsletters for your users without manual research.
• Automation engineers: maintain a repeatable n8n pipeline that links queue management (Google Sheets) to publishing (Gmail).
• Operators: batch-send newsletters to a subscriber list while tracking status in Google Sheets.

Technical details

• Integrations: Google Sheets (read/write), Gmail (send), OpenRouter (OpenAI models), Tavily (HTTP Header Auth).
• Workflow nodes: if, set, code, wait, gmail, merge.
• Key logic: topic selection (manual or Reddit via Tavily), hex ID creation, three-result research, three-section planning/writing, HTML conversion, and queued status updates.