> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bettertoken.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# How to connect n8n to BetterToken custom models

> Create an OpenAI credential in n8n with the BetterToken Base URL, API Key, and a GPT Key group model ID, then connect it to OpenAI Chat Model.

To connect n8n to BetterToken, use the built-in **OpenAI credential** and **OpenAI Chat Model**. Set the Base URL to `https://www.bettertoken.ai/v1`, use an API Key from the **GPT Key group**, and select the matching model from the model list.

## Prerequisites

* An n8n Cloud or self-hosted instance
* BetterToken API Key (<a href={"https://www.bettertoken.ai/register"}>register here</a>)
* A **GPT Key group model ID** copied from the <a href={"https://www.bettertoken.ai/pricing"}>model plaza</a>

| Setting         | Value                                  |
| --------------- | -------------------------------------- |
| Credential type | `OpenAI`                               |
| Base URL        | `https://www.bettertoken.ai/v1`        |
| API Key         | your BetterToken GPT Key group API Key |
| Organization ID | leave blank                            |
| Custom Header   | off                                    |

<Note>
  Keep `/v1` at the end of the Base URL. Do not add `/chat/completions` or `/responses`; n8n appends the route required by the node.
</Note>

## Setup

<Steps>
  <Step title="Add AI Agent and OpenAI Chat Model">
    Open a workflow and add an **AI Agent** node. At its **Chat Model** connector, add the **OpenAI Chat Model** sub-node.

    This guide uses **OpenAI Chat Model** because it supplies the model connection for n8n AI Agents, chains, and other AI workflows.
  </Step>

  <Step title="Create an OpenAI credential">
    In **OpenAI Chat Model**, open **Credential to connect with**, select **Create new credential**, and choose **OpenAI**.

    If your n8n version offers both **OpenAI Account (ChatGPT)** and **API Key**, choose **API Key**. OpenAI Account authentication is for an official account, not a BetterToken custom Base URL.
  </Step>

  <Step title="Enter the BetterToken credential">
    Use these values:

    | n8n field                  | Value                           |
    | -------------------------- | ------------------------------- |
    | API Key                    | your BetterToken API Key        |
    | Organization ID (optional) | leave blank                     |
    | Base URL                   | `https://www.bettertoken.ai/v1` |
    | Add Custom Header          | off                             |

    Click **Save**. n8n tests this credential against `/models` under the configured Base URL, so the complete test URL is `https://www.bettertoken.ai/v1/models`.
  </Step>

  <Step title="Select a model">
    Return to **OpenAI Chat Model**, open **Model**, and select the GPT Key group model ID shown in the <a href={"https://www.bettertoken.ai/pricing"}>model plaza</a>.

    If the list does not update after you save the credential, reopen the node or credential and load the model list again.
  </Step>

  <Step title="Choose Chat Completions or Responses">
    For the first test, keep **Use Responses API** off and complete a plain conversation through the default Chat Completions mode.

    Turn on **Use Responses API** only when your workflow needs it. n8n documents built-in tools such as Web Search, File Search, and Code Interpreter for **OpenAI Chat Model + AI Agent** with Responses enabled. Those OpenAI-hosted tools are not automatically capabilities confirmed for BetterToken, so test them against the selected model and endpoint.
  </Step>

  <Step title="Run a test">
    Give the AI Agent a simple input:

    ```text theme={null}
    Reply with exactly: Connection successful
    ```

    Click **Execute step** or run the workflow. The setup is complete when the node returns a model response.
  </Step>
</Steps>

## Troubleshooting

### Credential test returns 401

* Make sure the API Key is complete and has no extra spaces
* Confirm that the key belongs to the **GPT Key group**
* Check the key status and balance in the BetterToken Dashboard

### Credential test returns 404

* The Base URL must be `https://www.bettertoken.ai/v1`
* Do not use `https://www.bettertoken.ai/v1/models` as the Base URL
* Do not append `/chat/completions` or `/responses`

### The Model list is empty

* Save the credential again, then reopen **Model**
* Confirm that the API Key can access `https://www.bettertoken.ai/v1/models`
* Copy a current model ID from the <a href={"https://www.bettertoken.ai/pricing"}>model plaza</a>; do not use a retired or misspelled ID

### Plain chat works, but Agent tool calls fail

Turn off **Use Responses API** and all built-in tools, then verify plain Chat Completions. Enable Agent tools one at a time to identify whether the problem is the model capability, tool arguments, or workflow configuration.

### The request times out

Increase **Timeout** under **OpenAI Chat Model > Options** and keep a limited **Max Retries** value. Do not repeatedly retry `400`, `401`, or configuration errors.

## Scope

This guide confirms LLM Chat through the OpenAI-compatible protocol only. Do not configure a GPT chat model as an embedding, rerank, speech, or image model.

## Related docs

* [n8n OpenAI credentials](https://docs.n8n.io/integrations/builtin/credentials/openai/)
* [n8n OpenAI Chat Model](https://docs.n8n.io/integrations/builtin/cluster-nodes/sub-nodes/n8n-nodes-langchain.lmchatopenai/)
* [OpenAI-compatible API vs Anthropic-compatible API](/en/faq/concepts/openai-compatible-vs-anthropic-compatible)
* [How to choose a model](/en/faq/model-calling/model-selection-guide)