n8n AWS Bedrock Chat Model nodeConfigure AWS Bedrock Chat Model in n8n.
The n8n AWS Bedrock Chat Model node hands an Amazon Bedrock model to a root AI node. It exposes 3 parameters and two authentication methods, and it has no main input: it is a sub-node, a small node you clip onto the port of another node. Made for teams whose models already run inside AWS.
Verified Trustpilot reviews · AI, automation & growth agency
What does the n8n AWS Bedrock Chat Model node actually do?
It is the node that decides which large language model answers, when that model is served by Amazon Bedrock. n8n splits its AI nodes in two: a root node receives the workflow items through its main input, and sub-nodes clip onto its ports, one type of sub-node per port. This one is a sub-node. It never runs on its own, it never shows up alone in an execution, and it only exists to serve the node it is wired to.
Think of a support inbox where every message gets a draft reply. The root node is an AI Agent, and this node sits underneath it with Model pointing at a Bedrock model. Change the dropdown, the whole workflow changes model, and nothing else in the canvas moves. That is the practical reason the model lives in its own node.
Second scenario: an internal classifier. A workflow reads rows, sends each one to a root node for a verdict, and writes the label back to Google Sheets or posts the flagged ones to Slack. Here you want short, predictable output, so Sampling Temperature goes low and Maximum Number of Tokens stays small, because that field sets the length of the completion.
Third scenario: a company that already buys its compute on AWS. The credential is an IAM access key or an assumed IAM role, the traffic stays inside the AWS account, and billing stays on the AWS invoice. Bedrock charges its own API calls under its own terms, and n8n adds nothing on top of that.
When would you pick something else? If the model you want is not served by Bedrock, use the provider's own chat model node instead, for example the one that comes with OpenAI. Both plug into the same port, so the rest of the workflow is untouched. And if the question is whether a model is needed at all, remember that a plain workflow of regular nodes is cheaper and more predictable than any model call.
The known limits are worth stating up front. This node only offers 3 parameters, and the options collection carries two fields. There is no memory here, no tool, no output parser: those are other sub-nodes on other ports of the root node. The page describes node version 1, the maximum read in the catalog, and an older workflow can show fewer options. If your team is setting all this up for the first time, the n8n training walks through the root node and sub-node model, and the n8n review covers the platform itself.
What credential does Bedrock need in n8n?
- 01
Pick the authentication method
Open the node and look at Authentication. Two values are offered:
iam, labelled AWS (IAM), which uses an IAM access key, andassumeRole, labelled AWS (Assume Role), which temporarily assumes an IAM role. The second one suits organizations that already hand out roles rather than long-lived keys. The choice drives which credential type n8n asks for next. - 02
Create the AWS credential
With
iamselected, n8n asks for an AWS credential. WithassumeRole, it asks for an AWS (Assume Role) credential. A credential in n8n is a saved set of secrets: you fill it once from the Credentials menu and every workflow reuses it, so the key is never typed into a node. Keep one credential per AWS account rather than one per workflow. - 03
Check what the role is allowed to list
Back in the node, open the Model dropdown. It lists on-demand foundation models and cross-region inference profiles together. If part of that list is missing, the IAM role is usually short of the
bedrock:ListFoundationModelsorbedrock:ListInferenceProfilespermission. You can also switch the field to Expression mode and type a model ID, an inference profile ID, or an ARN directly.
Where does this node plug in?
A root node runs the logic and reads the workflow items; a sub-node like this one supplies a capability through a port, and a port only accepts one type of sub-node.
Output (what it plugs into)
- Model
ai_languageModel
01Model
This is the only wire the node has. It leaves the sub-node and lands on the Model port of a root node, which then sends every prompt through Bedrock and reads the completion back.
Key parameters
- Required: a root node without a chat model on this port has nothing to generate with, so the connection is the first thing to draw.
- AI Agent and Basic LLM Chain are the two usual destinations, the agent when tools are involved, the chain for a single prompt in and answer out.
- Text Classifier and Question and Answer Chain accept it too, for labelling items or answering from retrieved content.
What are the parameters of the AWS Bedrock Chat Model node?
The AWS Bedrock Chat Model node has 3 parameters. For each one: the node as you configure it in n8n, what the parameter changes, and our field notes.
Authentication
authenticationWhat you see in n8n
Notes & use cases
Sets how n8n proves to AWS that it may call Bedrock, and therefore which credential the node asks for. It is the first field to touch, before the model list can populate.
Key parameters
- AWS (IAM) (
iam): an IAM access key, the shortest path when a key already exists for the account. - AWS (Assume Role) (
assumeRole): the role is assumed temporarily, which fits accounts where long-lived keys are not handed out.
iam and moves to assumeRole when security tightens the rules.Model
modelWhat you see in n8n
Notes & use cases
Picks the model that generates the completion. Swapping this value is how a workflow changes model without any other edit on the canvas.
Key parameters
- Model: a dropdown listing on-demand foundation models and cross-region inference profiles together; the newest models are only available as inference profiles. Switch the field to Expression mode to pass a model ID, an inference profile ID, an ARN, or a value carried by the item, such as
{{ $json.model }}.
Options
optionsWhat you see in n8n
Notes & use cases
A collection, meaning an empty box you fill with Add Option. Nothing here is required, and leaving it closed is a perfectly reasonable default until an answer comes back too long or too loose.
Key parameters
- Maximum Number of Tokens: the ceiling on the tokens generated in the completion, so it is the field that caps answer length. A token is a chunk of text, roughly a short word.
- Sampling Temperature: controls randomness. Lower it and completions get less random; as it approaches zero the model turns deterministic and repetitive.
Need help automating AWS Bedrock Chat Model with n8n?
A person reads every message.
Questions builders ask about this node
01Is the AWS Bedrock Chat Model node included in n8n?
02What do you need to make the n8n AWS Bedrock Chat Model node work?
03What are the limits of this node?
04What is the minimum to plug in to get an answer?
05n8n or Make for a Bedrock workflow?
Get our weekly integration tips.
No spam. Unsubscribe anytime.
