Workflow Converter
Convert workflow JSON between n8n, Make, Zapier and Activepieces — with app mapping, confidence scoring, and an honest list of the steps that still need you.
Source workflow JSON
Target platform
Migrating? You may not need to rebuild it at all
Converting an existing workflow is the hard way round. Neura Market indexes 27,201 ready-made workflows for n8n, Make, Zapier, Activepieces and Pipedream — there is a good chance the automation you are porting already exists for your target platform, built natively for it. Import it and skip the conversion notes entirely.
Conversions, by direction
Every pair has its own quirks. Here is what carries across and what you will be rebuilding by hand.
n8n to Make
n8n nodes map to Make modules fairly cleanly for mainstream apps. The two differences that cause most rework are the data model — n8n passes an array of items per node, Make passes bundles — and n8n Code nodes, which have no direct Make equivalent.
Carries across
- Trigger type and schedule
- App actions where both platforms support the app
- Branching and filter conditions
- Field mappings and static values
Needs manual work
- Code nodes — rewrite as Make functions or split into modules
- Per-item iteration semantics — check whether an Iterator is needed
- Credentials — reconnect every connection in Make
- Self-hosted-only nodes with no Make counterpart
Make to n8n
The usual reason for this direction is cost or self-hosting. Make blueprints carry module configuration that maps well onto n8n nodes; routers become n8n branches, and Make functions usually become a single n8n Code node.
Carries across
- Module sequence and routing logic
- Trigger configuration
- Filters and conditions
- Most mainstream app operations
Needs manual work
- Make functions — consolidate into a Code node
- Iterator and Aggregator modules — n8n often needs neither
- Error handler routes — rebuild with n8n error workflows
- Credentials — reconnect in n8n
Zapier to n8n
The most common migration, usually driven by per-task pricing. Zapier Zaps are linear, so they translate to n8n readily. The gain is that multi-step branching and batch handling stop being awkward.
Carries across
- Trigger app and event
- Step order and app actions
- Field mappings
- Filter steps as n8n IF nodes
Needs manual work
- Zapier Paths — rebuild as n8n Switch or IF branches
- Formatter steps — usually a Set or Code node
- Credentials — reconnect in n8n
- Zapier-only built-ins with no direct equivalent
n8n to Zapier
The awkward direction, and the tool is explicit about why: Zapier publishes no importable workflow format. The converter therefore outputs a structured step-by-step specification to rebuild from, rather than a file to import.
Carries across
- A readable step specification with apps, actions and mappings
- Trigger definition
- Branch logic described as Zapier Paths
Needs manual work
- Everything — Zapier has no public JSON import, so the build is manual
- Per-item loops — Zapier runs one record per Zap run
- Code nodes — Zapier Code steps have a different runtime
n8n to Activepieces
Both are open-source-leaning and self-hostable, so the conceptual model transfers well. Coverage is the constraint: Activepieces has fewer pieces than n8n has nodes, so niche integrations may have no counterpart.
Carries across
- Trigger and step sequence
- Branching logic
- Core app actions where a piece exists
- Field mappings
Needs manual work
- Nodes with no matching Activepieces piece — replace with an HTTP step
- Code nodes — port to Activepieces code pieces
- Credentials — reconnect
What always carries across
- Workflow structure — the order of steps and how they branch
- App and action identity, where both platforms support the app
- Field mappings and static configuration values
- Trigger type, and schedules where both platforms express them the same way
What never does
- Credentials and API keys — never included in any export, by design
- Execution history and past run data
- Platform-specific features with no counterpart on the target
- Environment variables and instance-level settings
How to convert a workflow
- 1
Export the source workflow
In n8n, open the workflow and choose Download to get its JSON. In Make, use Export Blueprint. Zapier has no export, so start from a description of the Zap instead.
- 2
Paste it in and pick a target
The converter reads the structure — trigger, steps, branches, mappings — and identifies which app each step touches.
- 3
Review the confidence rating and notes
High confidence means most steps had direct equivalents. Low means the source leaned on platform-specific features. The notes name the steps needing attention rather than hiding them.
- 4
Import and reconnect
Import the output into the target platform, reconnect every credential, and run once against test data before enabling the trigger.
Frequently asked questions
How do I convert an n8n workflow to Make?+
Export the workflow from n8n (open it and choose Download to get the JSON), paste that JSON into the converter above, and select Make as the target. You get a Make-shaped blueprint plus a list of the steps that need manual attention — usually Code nodes and per-item iteration, which have no direct Make equivalent.
Can I convert a Zapier Zap to n8n?+
Yes, and it is the most common direction. Zaps are linear, so the step sequence maps onto n8n nodes readily. Zapier Paths become IF or Switch branches and Formatter steps usually become a Set or Code node. You will reconnect credentials in n8n afterwards — no export includes them.
Why does Zapier conversion output a text spec instead of JSON?+
Because Zapier does not publish an importable JSON format. Rather than emit a file that cannot be imported, the tool outputs a structured step-by-step specification you can follow in the Zapier editor.
Is the converted workflow guaranteed to work?+
No, and any tool claiming otherwise is overselling. Platforms differ in apps, capabilities and data models, so conversion is a high-quality starting point rather than a finished workflow. The confidence rating and conversion notes tell you which steps need review instead of leaving you to discover them at runtime.
What about credentials and authentication?+
Credentials are never copied — they are not present in any platform export, by design. The conversion preserves logic and app references; you reconnect each account in the target platform.
How large a workflow can I convert?+
Up to 30,000 characters of JSON, roughly 100-200 nodes depending on complexity. For larger workflows, split them into sub-workflows and convert each piece separately.
What is the confidence rating?+
How faithfully the source could be mapped to the target. High means most steps had direct equivalents. Medium means several approximations were made. Low means the source leaned on features unique to its platform and significant manual work will be required.
Is the workflow converter free?+
Yes — free to use with no signup and no account required.