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

# Refly

> Learn how to add an OpenAI-compatible provider in Refly, configure the Base URL, and use ScitiX models in AI workflow orchestration.

[Refly](https://github.com/refly-ai/refly) is an open-source AI workflow orchestration platform for vibe workflows. It supports custom **OpenAI**-compatible providers, so you can connect Refly to ScitiX Model Inference.

This guide uses `glm-5.2` as an example. For the full model list and pricing, see [Models](https://console.scitix.ai/model-inference/models).

## Configure

Create an API Key on the [API Keys](https://console.scitix.ai/model-inference/api_keys) page. After signing in to Refly, open **Settings > Model Config** or **Provider Config**, choose **Add Provider**, and fill in:

| Field    | Value                                                                                           |
| -------- | ----------------------------------------------------------------------------------------------- |
| Type     | **OpenAI**                                                                                      |
| Name     | A custom name, such as `scitix`                                                                 |
| Category | LLM                                                                                             |
| API Key  | Your API Key created on the [API Keys](https://console.scitix.ai/model-inference/api_keys) page |
| Base URL | `https://api.scitix.ai/model-api/v1`                                                            |

Choose **Test Connection**. When Refly shows **Connection test successful**, choose **Add**.

<img src="https://mintcdn.com/scitix-adfc65c2/p0x08-nbZl79pIrL/model-inference/media/refly-add-provider.png?fit=max&auto=format&n=p0x08-nbZl79pIrL&q=85&s=a567853ee736b69efc39d2575fcb00c0" alt="Refly OpenAI provider pointed to ScitiX with a successful connection test" style={{ width: "80%", height: "auto" }} width="3788" height="1718" data-path="model-inference/media/refly-add-provider.png" />

After saving the provider, choose **Add Model** and enter the model ID `glm-5.2`.

## Verify

Use the model in a conversation and send a message. A normal reply confirms that the connection works.

<img src="https://mintcdn.com/scitix-adfc65c2/p0x08-nbZl79pIrL/model-inference/media/refly-chat-reply.png?fit=max&auto=format&n=p0x08-nbZl79pIrL&q=85&s=22343de74bbfc3544cff267da90e484d" alt="Refly calls glm-5.2 through ScitiX and returns a normal reply" style={{ width: "60%", height: "auto" }} width="1496" height="1582" data-path="model-inference/media/refly-chat-reply.png" />

## FAQ

* **Provider Type**:
  Select **OpenAI**. Its Base URL is configurable, so no dedicated ScitiX preset is required. Refly uses LangChain `ChatOpenAI` in the backend.
* **Provider configuration location**:
  Providers are configured in the Web UI and saved to the account, not through environment variables. Use **Test Connection** after adding the provider to confirm connectivity.
* **Tool or workflow capabilities**:
  Choose a model that supports tool calling. `glm-5.2` supports tool calling.