Resources · n8n integration

n8n Item List Output Parser nodeConfigure Item List Output Parser in n8n.

One answer from a model, several items in the workflow. The n8n Item List Output Parser node splits what a chat model returns into separate items, using a separator and a maximum count. It carries 2 options, plugs into the Output Parser port of a root node, and never runs on its own.

Try the Item List Output Parser node on n8nSee the 1 parameters

Verified Trustpilot reviews · AI, automation & growth agency

Why automate

What does the n8n Item List Output Parser node do in a workflow?

A chat model answers in one block of text. When that answer is really a list, the rest of the workflow has to cut it up before anything useful happens. The n8n Item List Output Parser node does that cutting: it takes the raw answer of the root node it is attached to and returns the results as separate items, one per line by default. An output parser is a sub-node, meaning a small node you attach to a port of a bigger node instead of placing it in the main flow.

Say a model reads a long document and answers with study notes, one idea per line. Without a parser, the next node gets a single item holding the whole block, and a loop over it runs once. With this parser attached, each line becomes its own item, so a Google Sheets node appends one row per note instead of dumping a wall of text into a single cell.

Second case: a workflow reads a request for proposal and asks a model to pull out the questions it contains. Each question arrives as a separate item, which is exactly what a Slack node needs to post them one by one, or what a second model call needs to answer them one at a time.

Third case: an AI Agent that plans work. You ask it for the steps to take, the parser hands back the steps as items, and the branch after the agent runs once per step. Same idea with an OpenAI node further down the workflow that processes each item on its own.

When to reach for something else: this node returns a flat list of strings, nothing more. If the workflow needs typed fields, an object per result with a name, a date and an amount, splitting text on a separator is the wrong tool and you want a parser that returns structured JSON. And if the model answer is already a clean list inside a field of the item, a plain expression like {{ $json.text }} feeding a splitting node does the job without any AI sub-node.

The known limits are worth stating up front. The node has no main input and no settings beyond its 2 options, so it cannot clean the answer, deduplicate it, or repair malformed output. It also depends entirely on the model actually respecting the separator you asked for in the prompt: nothing here forces the format. Version 1 is the version described on this page, and an older workflow may show fewer options. The comparison with other platforms, and what self-hosting changes, sits in the n8n review.

Connections

What do you attach this sub-node to?

n8n splits AI work between a root node, which receives the items of the workflow, and sub-nodes attached to its ports, one type of sub-node per port.

Output (what it plugs into)

  • Output Parserai_outputParser
01Output ParserThis node has no main input. It exposes a single connection, the Output Parser port, and that port is how a root node reads its answer through the parser instead of raw.

This node has no main input. It exposes a single connection, the Output Parser port, and that port is how a root node reads its answer through the parser instead of raw.

Key parameters

  • Optional: a root node runs fine without any parser attached, and returns its answer as text.
  • AI Agent: the parser turns the agent final answer into one item per line, which the branch below then loops over.
  • Basic LLM Chain: the simplest place to start, a single prompt whose answer comes back already split.
  • Auto-fixing Output Parser: takes this node as its own parser and asks the model again when the answer does not parse.
Use cases
try it on a Basic LLM Chain with a short prompt, then move the same setup onto an AI Agent Tool.
Parameters

Which parameters can you set on the node?

The Item List Output Parser node has one parameter. For each one: the node as you configure it in n8n, what the parameter changes, and our field notes.

01

Options

options

What you see in n8n

Notes & use cases

Everything configurable sits behind Add Option. Untouched, the parser splits on a new line and returns every line it finds. The collection opens when the model answers in another shape.

Key parameters

  • Number Of Items: how many items are returned maximally. At -1 there is no limit, which suits an export; cap it when each item triggers a costly action.
  • Separator: the separator used to split the results into separate items. Defaults to a new line, and becomes a comma when the prompt asked for an inline list.
Use cases
keywords requested on one line come back comma separated, so Separator gets a comma and each keyword becomes its own item.
Need help

Need help automating Item List Output Parser with n8n?

A person reads every message.

FAQ

Item List Output Parser and n8n, the questions that come next

01Is the n8n Item List Output Parser node included in n8n, on Cloud and self-hosted?
Yes. The node ships with n8n as part of its AI package, and there is nothing to install and nothing extra to pay on the n8n side, whether the workflow runs on n8n Cloud, the hosted offer run by n8n, or on your own instance under the Community Edition and its Sustainable Use license. A workflow behaves the same in both places. The only billing involved in an AI workflow comes from the model provider, OpenAI, Anthropic, Google, Mistral or another, which charges its own API calls under its own terms. The parser itself calls no external service, so it adds nothing to that bill.
02What do you need to set up before the node works?
Nothing on this node. It has no credential and no Authentication selector, so there is no account to connect and no key to paste. What needs a credential is the chat model attached to the root node, since that is the sub-node talking to the provider. Practically: add a root node such as an AI Agent or a Basic LLM Chain, attach a chat model with its credential, then drag this parser onto the Output Parser port. From there the only thing left is the prompt, which should ask the model for one result per line so the default separator has something to cut on.
03What are the limits of this node?
It splits text, and that is all it does. The node returns a flat list of strings with no field names and no types, so an answer that should become an object with several properties is out of its reach. It has 2 options and no others: no trimming, no deduplication, no filtering of empty lines. It has no main input either, so it cannot be dropped into the flow between two regular nodes, and it never runs on its own. It also cannot force the model to respect the separator, that part belongs to the prompt. Version 1 is described here, and an older workflow may expose less.
04What is the minimum to plug in, and which sub-node should you start with?
The minimum is a root node plus a chat model, since a root node without a model cannot answer at all. This parser comes third and stays optional. To start, take a Basic LLM Chain, attach a chat model with a working credential, write a prompt that asks for one item per line, and attach this node to the Output Parser port. Run it once and check the output panel: you should see several items rather than one. Once that works, the same parser moves onto an AI Agent or an AI Agent Tool without any change to its settings, and an Auto-fixing Output Parser can wrap it if the model sometimes drifts from the format.
05n8n or Make for splitting a model answer into items?
It depends on where the workflow lives and how you want to pay for it. Make is a hosted platform, with no self-hosting option, billed per operation, which suits a team that wants nothing to run and a visual builder above all. n8n runs on Cloud or on your own servers with Docker or npm, so the data stays where you decide, which matters when the text sent to a model is a customer document. On the AI side, the root node and sub-node model is specific to n8n: the parser is a piece you attach and move between root nodes. Both handle the job, the choice sits on hosting, data control and cost model.
Hack'celeration Lab

Get our weekly integration tips.

No spam. Unsubscribe anytime.

Same platform

Other docs you might need on this stack.

Browse all integrations