Manifest now has a native n8n community node. The package is available on npm as n8n-nodes-manifest and ships two nodes: a
Manifest Chat Model sub-node that plugs Manifest into n8n AI Agents, and a
Manifest action node that calls the API directly from a workflow, without configuring a generic HTTP Request node.
Now live on n8n: n8n has reviewed and published the node as a verified community node, so it installs from
Settings > Community Nodes on n8n Cloud and on self-hosted n8n alike.
Use Manifest as the model behind an n8n AI Agent #
The Manifest Chat Model is a language-model sub-node built on n8n’s @n8n/ai-node-sdk. Connect it to the
AI Agent or
Basic LLM Chain node the same way you would connect the
OpenAIChat Model, and Manifest sits between the agent and the providers:
- The model list is loaded from your Manifest instance; the default
auto
lets Manifest route every request. - Streaming and native tool calling are supported, so AI Agent tools work as expected.
- An optional setting switches the sub-node to the OpenAI Responses API.
The sub-node requires an n8n version that ships @n8n/ai-node-sdk
(2026 releases).
Call Manifest directly from a workflow #
The Manifest action node covers workflows that call the API without an AI Agent. It includes three operations:
Create Chat Completion for OpenAI-compatible chat requestsCreate Response for OpenAI-compatible Responses API requestsList Models to inspect the models available to a Manifest agent key
Both nodes work with Manifest Cloud and self-hosted Manifest deployments.
Install the Manifest node in n8n #
In n8n, on Cloud or self-hosted:
- Open
Settings > Community Nodes. - Select
Install. - Enter
n8n-nodes-manifest
. - Confirm the community-node warning and install the package.
These steps follow n8n’s community-node installation guide.
Connect n8n to Manifest #
Both nodes share one Manifest API credential:
Base URL:https://app.manifest.build
for Manifest Cloud, or the root URL of your self-hosted deploymentAPI Key: the agent API key shown in Manifest
The Base URL must not end in /v1
; the nodes add the API path for each operation. If n8n runs in a container, a self-hosted Manifest URL must be reachable from that container.
Set Model to auto
to use the routing configured for the Manifest agent. For a first test, choose Create Chat Completion on the action node and use:
[
{
"role": "user",
"content": "Reply with one sentence confirming that Manifest is connected."
}
]
Then execute the node. The selected Manifest route handles the request and the node returns the response as n8n workflow data.
Routing owns the response mode #
Manifest routing determines whether a route returns buffered JSON or a streamed response. You do
not need to add stream
to Additional Body on the action node. For streamed routes, the node exposes the parsed server-sent events after the n8n step completes. The Chat Model sub-node streams natively into the AI Agent.
The node source lives in the Manifest repository. See the n8n setup page for the credential values and package links in one place.