N8n Merge Node Doubts

In the following 2 workflows 1 merge node gives output and other does not, why?? the inputs connected did not execute

Hi @Milyn_Dsilva_CC The Merge node only outputs once both of its inputs have actually run and sent data. In your second workflow one input never executes, so the Merge node has nothing to merge and returns no items.

Hi @Milyn_Dsilva_CC :waving_hand:

Additionally

If you don’t want the workflow to depend on both inputs, you should configure the Merge node with Mode: Append.
This is ideal when you just want to “merge whatever comes in”, even if only one input executes.

However, if your case requires the workflow to run only when BOTH inputs exist, the Merge node must require execution from both sides. In this case, use Combine, Merge by Key, or Wait.
If one input does not run or returns 0 items, no output will be produced.

1 Like

here its set to to merge (append) why did it not flow?

@Milyn_Dsilva_CC

Sorry for the confusion in my previous message
I didn’t explain it clearly in English.

According to the n8n documentation, the Merge node always waits for both inputs to execute, regardless of using Append. The mode only controls how the data is combined, not whether one input can be skipped. In Append, items from both inputs are simply concatenated, but the node still requires execution from both sides; if one branch doesn’t run, the Merge will not produce output.

In your case, if you’re using a recent version of n8n and you want the workflow to continue even when only one branch runs, you can set the Merge node to Pass-through, which forwards the data from a single selected input instead of waiting to combine both. Otherwise, the alternative is to redesign the workflow so both branches always execute, even if one of them only returns an empty item.

Then how does this merge node provide a output when one input is not executed

change switch to pass trough

1 Like

:smiling_face_with_tear::face_holding_back_tears::face_holding_back_tears::face_holding_back_tears::face_holding_back_tears: what switch, im sorry
please explain me why the merge node gives output here and not on next one

@Milyn_Dsilva_CC

Everything is fine, we will get there.
Go to the node merge parameters and change it to pass through

the version that im using does not have this, also both screenshots are append and not pass-through mode which i shared

@Milyn_Dsilva_CC

@Milyn_Dsilva_CC

I got a bit confused with the thread too, but I think I understand it now. The Merge node doesn’t require both inputs to have items, but it does require both branches to execute. One input can have 0 items and the Merge will still run, but if a branch never executes, the Merge won’t produce output.

Could you please confirm in the execution view whether both Input 1 and Input 2 actually executed (turned green)? That will help confirm what is happening in this case.

@Milyn_Dsilva_CC The node you are using as the second input in your example is a ‘No Operation’ node which means it does nothing, and when it does nothing which means it would not do anything and that is why it get accepted as there would never be a input coming from that side, and in your main workflow the merge does not work because the ‘Opportunity API’ and ‘Custom field API’ never returns any error which should be required to trigger the second branch of your workflow, that is why your merge does not work, let me know if you need more clarification.

1 Like

Hey @Milyn_Dsilva_CC Welcome to the n8n community !

I agree with @Anshul_Namdev If the branch connected in the second input of the merge node doesnt get executed then the merge node is going to output the value in input 1. The merge node can be confusing and it has multiple functionalities accross different usecases. I have created a small workflow for you to show the functionality of the merge node clearly.

In your usecase if you want it to wait for both outputs and assuming both branches are getting executed you can use the “combine“ option to combine the inputs of the 2 branches.

1 Like

I hope this helps you @Milyn_Dsilva_CC

1 Like

Thanks @Deepankar @tamy.santos @Anshul_Namdev

2 Likes

@Milyn_Dsilva_CC You’re welcome! We hope you found the solution. If so, please mark the reply that best answers your question to help future users and support the community.

Appreciate it @Milyn_Dsilva_CC Happy flowgramming !

I would appreciate if you can mark the reply which helped you as a solution so that it helps other people who face the same issue in the future as well and it also gives us a nice push to contribute more.

Thanks