Descriptions:
Nate Herk presents a step-by-step method for migrating large numbers of n8n workflows between instances — specifically from n8n Cloud to a self-hosted deployment — using n8n’s own API and Google Sheets as an external tracking layer. The solution is demonstrated live against a real cloud account containing 512 workflows.
The migration system uses two separate n8n workflows. The first pulls all workflows from the source instance via the n8n node, loops through them, and writes each one as a JSON record to a Google Sheet, capturing metadata like creation date, active status, node configurations, and connections. This logging step prevents accidental duplicate imports and provides a checkpoint if something fails mid-migration. The second workflow reads back rows from the sheet filtered by a status column (set to “process” to indicate ready-to-migrate items), fetches each workflow’s full JSON by ID from the source instance, and creates it in the target instance using the destination API key.
Herk covers credential setup for both the source (n8n Cloud) and target (self-hosted) instances, including how to construct the correct API base URL format (base-url/api/v1), and demonstrates limiting the test run to 25 workflows to keep the recording manageable. The complete workflow is available as a free download. Herk also mentions using Hostinger VPS for self-hosting and references his AI Plus community for additional discounts.
📺 Source: Nate Herk | AI Automation · Published January 14, 2026
🏷️ Format: Workflow Case Study
