Should you upgrade to n8n 3.0 right now? Not on your production instance. n8n announced 3.0 as a beta on 27 October 2026 and says the stable release follows in early December. It is a major release with breaking changes, so the safe path is: run the built-in Migration Report, test on a copy, take a full backup, and only then upgrade. This guide covers what that means for workflows that call scrapers and Apify actors.
What is n8n 3.0 and when does it become stable?
According to n8n's own announcement, n8n 3.0.0 entered beta on 27 October 2026, with the stable 3.x line planned for early December 2026. n8n follows semantic versioning, so anything that can force a change to your workflows or configuration waits for a major version. 3.0 bundles those changes: legacy code paths removed, safer defaults, and a few features retired.
Third-party write-ups disagree on small details (one older docs page still described an October release), so treat the official v3.0 breaking changes page as the final word and re-check it before you act. Everything below is what those sources reported when this post was written, not something we tested on our own instance.
What breaks in n8n 3.0?
These are the changes the official docs and the guides summarising them list. Check each against your own setup.
| Change | Who it hits | What to do |
|---|---|---|
| Self-hosting is Docker-based; npm / npx installs are reported as unsupported | Anyone running npx n8n or a global npm install on a VPS | Move to the official Docker image before upgrading |
| Some legacy nodes are removed, including the Function and Function Item nodes | Older workflows, including copied templates | Replace them with the Code node |
| Code node timeout drops from 5 minutes to 1 minute | Code nodes that parse or reshape big scraper datasets | Split the work, or raise N8N_RUNNERS_TASK_TIMEOUT |
| Chat hub and non-functional nodes are retired or removed | Workflows that reference them | Find them in the Migration Report |
| Community node development needs Docker or Podman | Teams that build their own nodes | Update the dev environment |
The upgrade is also one-way. Going back means restoring the full backup you took before, so the backup is not optional.
How do I check my workflows before upgrading?
n8n ships a Migration Report in Settings > Migration Report, available from version 2.39. It scans your instance and lists the workflows and instance settings that 3.0 changes. n8n's docs warn that an empty report does not mean zero risk: the breaking changes page lists changes the tool cannot detect.
- Update to a 2.x release of 2.39 or later.
- Open the Migration Report and export or screenshot the list.
- Read the full v3.0 breaking changes page for the items the tool cannot see.
- Spin up a copy of your instance from a backup, upgrade the copy, and run your real workflows on it.
- Upgrade production only after the stable release, in a quiet window.
Which scraping workflows are most exposed?
A typical scraping automation is a schedule trigger, an Apify run, a Code or Set node that tidies the output, then Google Sheets, Telegram or Slack. Two spots deserve attention.
1. The tidy-up Code node. If it loops over thousands of rows, the shorter 1-minute Code node timeout is the likeliest thing to bite. The fix is easier on the scraping side: ask the actor for fewer, cleaner fields, or cap the run size, so the Code node has less to chew on. Most Apify actors let you set a maximum number of items per run, for example the Wallapop Scraper takes a Max items setting and has a monitor mode that returns only new listings and price changes. Smaller, incremental runs mean small payloads in n8n.
2. Old nodes inherited from copied workflows. Workflows pasted from older tutorials often use Function or Function Item nodes. Search your instance for them now, while nothing is broken.
If you use a ready-made workflow, check that it is built on current nodes. Our Wallapop alerts to Telegram and Google Sheets template is one of the n8n templates we publish; the workflow file is on the page, so you can open it on your test copy and see how it behaves after the upgrade.
Is it safe to wait?
For most teams, yes. The 2.x line was still receiving fixes in early November per n8n's release notes, and the stable 3.x line is planned for December. Waiting costs you the new features; rushing risks a one-way upgrade on workflows you have not tested. If a workflow earns or saves you money, wait for stable and test on a copy first.
Does this change how I use Apify actors from n8n?
The Apify side does not change: an actor is still called through its API or the Apify node and returns a dataset. What changes is how robust the n8n side must be. A few habits make any upgrade less painful:
- Keep the heavy lifting (pagination, proxies, dedupe) in the actor, not in n8n Code nodes.
- Schedule runs on Apify and use webhooks, so a broken n8n workflow does not lose data.
- Keep a short list of which workflows depend on which actors and which nodes.
For price and listing monitors, see how to track competitor prices automatically. If you are still deciding on a tool, read n8n vs Zapier vs Make for AI workflow automation. Browse everything we publish for marketplaces under E-commerce & Marketplace Scrapers.
What has n8n shipped in 2.x that you get anyway?
You do not need 3.0 to use n8n's recent AI features. Per n8n's release notes, an AI Assistant for self-hosted instances arrived in August 2026, and release 2.36 on 2 September 2026 added the option to use AI models without setting up separate provider accounts. A later 2.46.1 release added a permission check to the Assistant and AI-builder flows. If the AI features are what you want, you can have them on a current 2.x release while 3.0 settles.
A pre-upgrade checklist for scraping workflows
Use this list on your test copy. It is built from the changes above, not from a benchmark we ran.
- Inventory: list every workflow that calls an Apify actor or any scraper, and note the nodes after the call.
- Find removed nodes: search for Function and Function Item nodes and replace them with Code nodes, then re-run.
- Time your Code nodes: run each against a realistic dataset. Anything near a minute needs splitting or a higher
N8N_RUNNERS_TASK_TIMEOUT. - Check how n8n is installed: if it runs from npm, plan the Docker move first, as a separate step.
- Test the full path: trigger a scheduled run and confirm rows reach the Sheet, and alerts reach Telegram or Slack.
- Keep the backup: store the pre-upgrade backup somewhere you can restore from, because rollback depends on it.
Why keep scraping logic out of n8n?
Any time the workflow engine changes, logic that lives inside it is exposed. Logic that lives in an actor is not. An actor handles pagination, proxies and retries on its own schedule, and it can run on Apify even while your n8n instance is down for an upgrade. When n8n is back, it only has to read a finished dataset. That separation is the main reason we publish the n8n templates as thin workflows around an actor: the template reads results and routes them, and the actor does the scraping.
Frequently asked questions
n8n announced the 3.0.0 beta on 27 October 2026 and says the stable 3.x release is planned for early December 2026.
Workflows that use removed nodes, such as Function and Function Item, are reported to fail, and Code nodes that run longer than the new 1-minute default timeout may fail too. Run the Migration Report and test on a copy to find out which of yours are affected.
Not in place. The upgrade is described as one-way, so a rollback means restoring the full backup you took before upgrading.
No change on the Apify side is documented. The risk sits in the n8n nodes around the actor call, so check those with the Migration Report.
Want fresh marketplace data flowing into your workflows without heavy Code nodes? Try our actors on Apify. New accounts get free credit to test.
Sources: n8n announcement blog for 3.0 beta, the n8n docs v3.0 breaking changes and migration tool pages, and the n8n community announcement thread. Not tested by us on a live instance.
Last updated: November 2026
โก Run this without building it
These actors already do what this guide describes. Free $5 credit on a new Apify account.
๐ Keep reading
Get the Free Web Scraping Toolkit
Join the newsletter and get my curated list of scraping tools, proxy comparison cheatsheet, and Python automation templates.