We are facing a recurring issue when working with workflows that include the Telegram Trigger node.
Problem description
When activating a workflow that contains a Telegram Trigger, the trigger does not start listening to Telegram messages right after activation.
This happens both when activating the workflow via API and manually through the UI.
There are no error messages in execution logs or in the browser console.
The only way to make the trigger functional is to edit any node in the workflow and save it again. After doing this, the Telegram Trigger starts working as expected.
Steps to reproduce
Create a workflow with a Telegram Trigger node.
Activate the workflow (through the UI or the REST API).
Send a message to the Telegram bot.
Expected result: the workflow should be triggered.
Actual result: nothing happens until you manually edit a node and save the workflow again.
Additional information
The issue is consistently reproducible (not intermittent).
We already confirmed that:
The bot token is valid.
No other workflows are using the same bot.
HTTPS endpoint is valid.
No error logs are generated.
Environment
n8n version: 1.91.2
Deployment: [Docker/Server/Other – specify your setup if needed]
Question:
Has anyone experienced this behavior with Telegram Trigger?
Are there known workarounds (e.g., reusing webhookId, forcing webhook re-registration, or additional activation steps) to make the trigger work immediately after activation?
Hey, can you let me know if you are using more than one Telegram trigger across workflows? I think Telegram allows one per bot, so it will overwrite the old ones with the latest one that gets registered.
Also, could you try creating a new workflow with just a telegram trigger, disable any others for the same bot and test there?
Yes, I do have another workflow with a Telegram Trigger, but that one is currently inactive. The issue is that the trigger in my active workflow only starts working properly after I move any node and re-save the workflow. Once I do that, it begins responding as expected, but not before.
I already contacted support, and they suggested updating my n8n version. However, I’m hesitant because I’m concerned that updating might break my current setup, since I rely heavily on the API and don’t want to risk compatibility issues.
To be fair your current version is a bit outdated and I would have also suggested updating at some point. You could exporting your workflows and credentials and installing a separate instance of n8n on a newer version, then importing them there.
You mentioned you rely on the API - is this the n8n api? Because if so, you might find this important:
Yes, I use the n8n API, but the issue is also with the n8n UI. I’m hesitant because I use the API to generate workflows. If I update and break my code, can I roll back with no problem?
You can certainly roll back, however, I can’t guarantee no problems It’s dependent on a lot of things and the more versions you skip the more potential issues.
That’s why I recommend installing a separate instance if possible, on the new version where you can at least test your trigger. Also, you can comb through the changelogs with AI to make sure there aren’t breaking changes related to the API endpoints.
It’s dependent on a lot of things and the more versions you skip the more potential issues.
