AI Agent n8n: Podcast to Newsletter Auto

The Manual Trigger node serves as the entry point for your podcast-to-newsletter automation. This node allows you to initiate the workflow on-demand by clicking the "Execute" button directly in the n8n canvas.

In a production environment, you might replace this with a Schedule Trigger to run automatically at specific times, or a Webhook Trigger to start when new files are uploaded. For testing and initial setup, the Manual Trigger gives you complete control over when the workflow executes.

💡 Tip: Once your workflow is tested and working, consider replacing this with a Schedule Trigger set to run daily or weekly, automatically processing any new podcast files in your folder.

This Google Drive node searches your designated folder for audio files that need to be processed. It's configured to look specifically within your audio folder and return all matching files, making it easy to batch process multiple podcast episodes.

The search operation scans your Google Drive and retrieves file metadata including the unique file IDs needed for the download step. By filtering to a specific folder, you ensure the workflow only processes files you've explicitly placed there for newsletter conversion.

💡 Tip: Create a dedicated folder in Google Drive specifically for podcast files awaiting processing. Once the workflow runs, you can move processed files to an "archive" folder to avoid reprocessing.

After identifying the audio files, this node downloads the actual file content from Google Drive. It uses the file ID passed from the previous search step to fetch the correct audio file and prepare it for transcription.

The downloaded file is stored as binary data within n8n, which is then passed to the OpenAI transcription node. This approach handles files of any supported format (MP3, M4A, WAV, etc.) without requiring manual intervention.

💡 Tip: Ensure your audio files are under 25 MB for OpenAI transcription. For longer podcasts, consider splitting episodes into segments before uploading, or use a compression tool to reduce file size.

This OpenAI node leverages the Whisper model to transcribe your podcast audio into text. Whisper provides highly accurate transcriptions across multiple languages and handles various audio qualities well, making it ideal for podcast content.

The transcription output becomes the foundation for your newsletter. The AI agent in the next step will analyze this text to extract key insights, structure the content, and generate the final newsletter format. Learn more about leveraging OpenAI with n8n for powerful automations.

💡 Tip: The Whisper model supports multiple languages and can auto-detect the spoken language. If your podcast is multilingual, the transcription will handle language switches naturally.

The AI Agent node is the brain of this automation. It receives the podcast transcript and generates a complete HTML newsletter with a compelling subject line. The node is configured with strict output formatting rules to ensure the generated content can be reliably parsed and sent.

This configuration uses a system message that enforces JSON output containing exactly two fields: "subject" for the email subject line and "html" for the complete newsletter body. This structured approach eliminates parsing errors and ensures consistent output quality.

💡 Tip: The system message is crucial for reliable output. If you modify it, always test thoroughly to ensure the AI still produces valid JSON that downstream nodes can process.

This node configures the specific OpenAI model used by the AI Agent. GPT-4.1-mini provides an excellent balance between quality and cost for newsletter generation tasks—it's capable enough to produce engaging content while keeping API costs reasonable for regular use.

The Responses API option is enabled to improve interaction handling and provide better structured outputs. This setting works in conjunction with the output format requirements defined in the AI Agent node.

💡 Tip: If you need higher quality output for premium newsletters, you can switch to gpt-4 or gpt-4-turbo. For cost-sensitive operations with simpler content, gpt-3.5-turbo is also a viable option.

This JSON node creates a schema that defines the expected structure for newsletter output. By providing an example JSON, you establish the contract that the AI must follow when generating content. This ensures consistency across all generated newsletters.

The schema specifies two required properties: a subject line for the email and the complete HTML content. The Auto-Fix Format option helps correct minor formatting issues automatically.

💡 Tip: Customize the JSON example to match your newsletter style. Include a realistic subject line example that reflects your brand voice, and a placeholder indicating the full HTML structure you expect.

This additional AI Agent configuration ensures the model connection is properly established with your OpenAI credentials. It mirrors the ChatGPT node settings to maintain consistency throughout the newsletter generation process.

The configuration uses the same GPT-4.1-mini model with the Responses API enabled, ensuring all AI interactions follow the same parameters and produce consistent results.

This Google Sheets node retrieves your newsletter subscriber list from a spreadsheet. It's configured to fetch all rows from your subscriber sheet, providing the email addresses needed for the final email distribution step.

Using Google Sheets as your subscriber database offers flexibility—you can easily add, remove, or segment subscribers without modifying the workflow. The node pulls the current list each time the workflow runs, ensuring you always send to the most up-to-date audience.

💡 Tip: Structure your Google Sheet with clear column headers like "Email", "First Name", "Subscribed Date". This makes it easy to personalize newsletters and segment your audience in the future.

The SplitInBatches node divides your subscriber list into individual items for processing. With a batch size of 1, each subscriber is processed separately, ensuring reliable email delivery and easier error handling if any individual send fails.

This approach also helps respect rate limits imposed by email service providers. By processing one subscriber at a time, you avoid overwhelming the Postmark API and ensure each email is properly tracked.

💡 Tip: If you have a large subscriber list and want faster processing, you can increase the batch size. However, keep it reasonable (5-10) to avoid rate limiting issues with your email provider.

This Set node consolidates all the necessary data for email sending. It maps the generated newsletter subject and HTML content from the AI Agent output, along with the current subscriber's email address, into a clean data structure.

By explicitly setting these fields, you ensure that downstream nodes have reliable access to the exact data they need, regardless of any changes in upstream node outputs.