---
updatedAt: 2026-09-29T19:11:12.000Z
agentTools:
  projectIndex: https://docs.ascerta.com/llms.txt
---

# Configuring n8n Workflows for Ascerta

# Overview

The nodes in `@ascerta/n8n-nodes-ascerta` route model requests through the Ascerta proxy to the selected upstream provider. The five chat-model nodes connect to an n8n AI Agent; the standalone **Ascerta Proxy** node can make provider requests in a regular workflow.

## Tracking fields

All Ascerta nodes expose these fields:

| Field         | Purpose and default                                                                                                   |
| :------------ | :-------------------------------------------------------------------------------------------------------------------- |
| User ID       | Optional user identifier for cost attribution.                                                                        |
| Use Case Name | Defaults to the n8n workflow name, with spaces replaced by hyphens.                                                   |
| Use Case ID   | Defaults to the n8n execution ID, so each run has its own use case instance.                                          |
| Use Case Step | Defaults to the node's display name; chat-model nodes use the canvas node name when the default has not been changed. |

Under **Advanced Tracking**, you can set **Use Case Version**, **Use Case Properties** (a JSON object), and **Limit IDs** (a comma-separated list of Ascerta limits to enforce). **Flatten Gemini Content** is enabled by default for Databricks schema compatibility. **Debug Logging** is disabled by default and can log sensitive request and credential details to the n8n server console; enable it only with guidance from Ascerta Support.

See [Use Cases](https://docs.ascerta.com/docs/use-cases), [user attribution](https://docs.ascerta.com/docs/day-2-tracking-users), and [Limits](https://docs.ascerta.com/docs/limits-continued) for Ascerta-side configuration.

## AI Agent chat-model nodes

Each chat-model node needs an **Ascerta API** credential plus its provider credential. The type IDs below are the fully qualified values used in n8n workflow JSON.

| Provider node                    | Type ID                                              | Provider credential    | Ascerta proxy path           |
| :------------------------------- | :--------------------------------------------------- | :--------------------- | :--------------------------- |
| Ascerta OpenAI (Proxy)           | `@ascerta/n8n-nodes-ascerta.lmChatAscerta`           | `openAiApi`            | `/api/v1/proxy/openai/v1`    |
| Ascerta Anthropic (Proxy)        | `@ascerta/n8n-nodes-ascerta.lmChatAscertaAnthropic`  | `anthropicApi`         | `/api/v1/proxy/anthropic`    |
| Ascerta Azure AI Foundry (Proxy) | `@ascerta/n8n-nodes-ascerta.lmChatAscertaAzure`      | `azureOpenAiApi`       | `/api/v1/proxy/azure.openai` |
| Ascerta Amazon Bedrock (Proxy)   | `@ascerta/n8n-nodes-ascerta.lmChatAscertaBedrock`    | `aws`                  | `/api/v1/proxy/aws.bedrock`  |
| Ascerta Databricks (Proxy)       | `@ascerta/n8n-nodes-ascerta.lmChatAscertaDatabricks` | `ascertaDatabricksApi` | `/api/v1/proxy/openai/v1`    |

The Ascerta credential type is `ascertaApi`. Both `ascertaApi` and `ascertaDatabricksApi` are included in this package; you do not need `n8n-nodes-databricks` for the Ascerta Databricks node.

### OpenAI and Anthropic

Select the corresponding provider credential and a model supported by that provider. The node forwards requests through Ascerta using the OpenAI or Anthropic proxy path shown above.

### Azure AI Foundry

This node is for Azure OpenAI deployments using the OpenAI-compatible API. Set **Deployment Name** to the Azure deployment name. The node uses the Azure credential's explicit endpoint if present; otherwise it constructs `https://<resourceName>.openai.azure.com`. **API Version** comes from the node field, then the credential, then `2024-08-01-preview`.

### Amazon Bedrock

Set **Model ID** to a Bedrock model ID available to your account. The node's **AWS Region** field defaults to `us-east-1` and overrides the region in the AWS credential. Ascerta routes requests through the Bedrock proxy with AWS signing.

### Databricks

Create an **Ascerta Databricks API** credential with a workspace URL and personal access token. Select a **Serving Endpoint Name** from the list or enter it by name, then choose the **Cloud Provider** (AWS, Azure, or Google Cloud). For a custom serving endpoint whose name does not start with `databricks-`, **Deployed Model** is required. You can select it from the list or enter it by name.

## Standalone Ascerta Proxy node

Use `@ascerta/n8n-nodes-ascerta.ascerta` when you need a normal workflow node instead of an AI Agent chat-model connection. Select **Provider** (OpenAI, Anthropic, Azure OpenAI, AWS Bedrock, or Databricks) and configure the provider-specific fields. It uses the **Ascerta API** credential, and Databricks requests also use **Ascerta Databricks API**. The node can include cost data and return the full response through its output options.