Production Webhook returns “Workflow could not be started!” despite active workflow

Environment

  • n8n Cloud

  • Workflow is marked as Published / Active

  • Webhook node version: 2.1

  • Using Production URL (not test URL)


Problem Description

I am experiencing an issue where a Production Webhook in n8n Cloud consistently returns:

{"code":0,"message":"Workflow Webhook Error: Workflow could not be started!"}

even though:

  • The workflow is published

  • The webhook path is unique (no duplicate paths in other workflows)

  • The Production URL is being used (not webhook-test)

  • The workflow contains only:

    • 1 Webhook node (POST)

    • 1 Respond to Webhook node

  • The Webhook node “Respond” parameter is correctly set to:
    Using ‘Respond to Webhook’ Node

  • The Respond to Webhook node is properly connected and configured


Webhook Configuration

HTTP Method: POST
Path: rfq/upload/submit-ok-1-v3
Authentication: None
Respond mode: Using “Respond to Webhook” node

The Respond to Webhook node is configured as:

  • Respond With: Text

  • Response Body: ok-post


How It Is Triggered

The webhook is triggered by a standard HTML form submission:

<form action="https://speedx.app.n8n.cloud/webhook/rfq/upload/submit-ok-1-v3"
      method="POST"
      enctype="multipart/form-data">

The form is submitted directly from a browser (normal POST form submission).


What Works

  1. A GET webhook test endpoint works correctly.

  2. Using browser console fetch() from within n8n UI also works and returns:

    ok-post
    
    
  3. If I remove the Respond to Webhook node and set Webhook to respond “Immediately”, I sometimes get a response.


What Fails

When submitting the form normally from the browser to the Production URL, I consistently receive:

{"code":0,"message":"Workflow Webhook Error: Workflow could not be started!"}

This happens even when:

  • The workflow is newly created

  • The webhook path is brand new and unique

  • There are no other workflows using the same path

  • The workflow contains only the Webhook and Respond nodes


Additional Observations

  • GET requests to POST-only endpoints correctly return 404 (as expected).

  • No duplicate webhook paths exist.

  • The workflow appears active.

  • No execution appears in the Executions tab when this error occurs.


Expected Behavior

Submitting a POST request to the Production Webhook URL should trigger the workflow and return the configured response (ok-post).


Actual Behavior

The workflow does not start and returns:

Workflow Webhook Error: Workflow could not be started!


If needed, I can provide:

  • Workflow export JSON

  • Screenshot of configuration

  • Exact webhook ID

  • Instance ID

Hi @Krystian_Dulawa
the issue is specifically multipart/form-data combined with the Respond to Webhook node. n8n tries to hold the connection open while parsing the heavy payload and spinning up a worker, and it just times out or crashes before the workflow technically starts.

I’d go with this option to fix it:

Stop using <form> tags: (best way) Use a simple JS fetch() script to send the data instead. Since you already saw fetch works, just stick with that. It handles the data stream way better than a raw browser navigation.
let me know if this works for you.