Overview

Each Pay-i Node routes requests through the Pay-i proxy to a specific upstream provider. Requests and the responses are passed through transparently, and un-modified. This allows Pay-i to be your system of record for AI usage, while ensuring that proper attribution is captured without modifying the payload.

Common Properties

Every Pay-i node exposes the following properties within it's invidivual UI. You can set these to associate requests with Pay-i Use Cases, users, and Limits. These fields are the same across all of the providers included within the n8n-nodes-payipackage.

Property	Description
Use Case Name	Associates each workflow execution with a Pay-i Use Case. If you do not specify this, the Pay-i node within n8n will use the workflow name
User ID	Per-user attribution within the Pay-i dashboard
Use Case Step	The step name from the n8n workflow canvas. If not defined, Pay-i will use whatever that step is called as it appears on your n8n canvas
Account Name	Per-account attribution within the Pay-i dashboard
Limit IDs	IDs of the Pay-i Limits - Currently this only supports reading limits
Properties	Custom key-value metadata for filtering and reporting within Pay-i

To learn more on how Use Cases and Limits work, see Use Cases and Limits.

AI Providers Supported within n8n

Every Pay-i node will forward AI provider prompt properties (model name, temperature, etc) as-is.

OpenAI


Node	`Pay-i OpenAI Proxy / lmChatPayi`
Credentials	`PayiApi + openAIApi`
Proxy Path	`/api/v1/proxy/openai/v1`
LangChain Class	`ChatOpenAI`
Pay-i Supported Models	See: `Category: system.openai`(Link)

Anthropic


Node	`Pay-i Anthriopic (Proxy) / lmChatPayi`
Credentials	`PayiApi + openAIApiAnthropic`
Proxy Path	`/api/v1/proxy/anthropic`
LangChain Class	`ChatAnthropic`
Pay-i Supported Models	See: `Category: system.anthropic`(Link)

Azure AI Foundry


Node	`Pay-i Azure AI Foundry (Proxy) / lmChatPayiAzure`
Credentials	`PayiApi + azureOpenAiApi`
Proxy Path	`/api/v1/proxy/azure.openai`
LangChain Class	`ChatOpenAI` (Not `AzureChatOpenAI` will avoid authentication conflicts with the proxy.
Pay-i Supported Models	See: `Category: system.azure`(Link)

Scope: This node targets Azure OpenAI endpoints - the OpenAI comptable API surgace within Azure AI Foundry., You can use other models, but this node is specifically designed to handle Azure OpenAI deployments using the OpenAI API wire format.
Required You must set the Deployment Name field to match your Azure deployment exactly.
Endpoint Resolution If your credentials have an explicit endpointvalue, use that. Otherwise the endpoint is assembled using resourceName as https://_[resourceName]_.openai.azure.com. The resolved endpoint will be sent as xProxy-Provider-BaseURI.
API Version Falls back in order: node parameter --> Credential value --> default (2024-08-01-preview)

For additional information and Azure OpenAI specific guidance, see Instrumentation for Microsoft Foundry and Azure Open AI contact Pay-i Support if you need assitance configuring Provisioned Throughput Units.

AWS Bedrock


Node	`Pay-i Anthriopic (Proxy) / lmChatPayiBedrock`
Credentials	`PayiApi + aws`
Proxy Path	`/api/v1/proxy/aws.bedrock`
LangChain Class	`ChatBedRockConverse`
Pay-i Supported Models	See: `Category: system.aws.bedrock`(Link)

Model Validation When using AWS Bedrock, you must set your Model to a validated Bedrock model ID. (for example, us.anthropic.claude-sonnet-4-6)
Setting your Region The region defaults from the n8n credentials you set. If no region is specified, or Pay-i doesn't know how to hanlde it, Pay-i will default to us-east-1
Signing Limitation There are some considerations for code that uses the AWS sigV4 function. Please contact [email protected] if you experience issues or need guidance prior to deploying Bedrock services into production.

Databricks


Node	`Pay-i Databricks (Proxy) / lmChatPayiDatabricks`
Credentials	`PayiApi + databricks` (Uses `n8n-nodes-databricks`)
Proxy Path	`/api/v1/proxy/openai/v1` ( Reuses the OpenAI path)
LangChain Class	`ChatOpenAI`
Pay-i Supported Models	See: `Category: system.databricks.[Hyperscaler]`(Link)

Cloud Provider Option Required You must select the Cloud Provider within the Pay-i node if you are using Databricks. AWS and Google use [hostname].cloud.databricks.com as their endpoint but Azure's Managed Databricks Service uses [hostname].azuredatabricks.net. If this value is not set, your data will not show up correctly or you may find random errors.
Hyperscaler Support Pay-i currently supports AWS, Azure, and Google Cloud as providers for Databricks. Contact Pay-i support if you need other options.