> ## 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.

# Flowise

> Learn how to configure Flowise ChatOpenAI Custom nodes with ScitiX Model Inference for low-code LLM workflows.

[Flowise](https://github.com/FlowiseAI/Flowise) is a popular low-code builder for LLM apps and agents. Its **ChatOpenAI Custom** node takes a custom `BasePath`, so flows can run on the ScitiX Model Inference API.

Default model here is `glm-5.2`; see the [Models](https://console.scitix.ai/model-inference/models) page for the full list and pricing.

Deploy Flowise per the [official guide](https://docs.flowiseai.com) and open the UI (default `http://localhost:3000`).

## Configure

Create an API key on the ScitiX [API Keys](https://console.scitix.ai/model-inference/api_keys) page. In Flowise, open **Credentials → Add Credential → OpenAI API**, name it (e.g. `Scitix`), and paste the ScitiX key.

In a flow, add a **ChatOpenAI Custom** node and set:

| Field                                      | Value                                                             |
| ------------------------------------------ | ----------------------------------------------------------------- |
| Connect Credential                         | The `Scitix` credential above                                     |
| Model Name                                 | `glm-5.2`. Pick a model with native tool calling for Agent nodes. |
| BasePath (under **Additional Parameters**) | `https://api.scitix.ai/model-api/v1`                              |

Use the **ChatOpenAI Custom** node, not plain **ChatOpenAI**. It exposes the `BasePath` field and does not validate the model name against OpenAI's own list, which is what lets ScitiX IDs like `glm-5.2` through.

<img src="https://mintcdn.com/scitix-adfc65c2/AblxxMM8pLmVYmdt/model-inference/media/flowise-basepath.png?fit=max&auto=format&n=AblxxMM8pLmVYmdt&q=85&s=f2980805fba214222fbedae67e936d99" alt="Flowise — ChatOpenAI Custom BasePath pointed at ScitiX" width="1262" height="1118" data-path="model-inference/media/flowise-basepath.png" />

## Verify

A Chatflow's **ending node must be a Chain / Agent / Engine**, not the model node itself. Add a **Conversation Chain**, wire the **ChatOpenAI Custom** node into its **Chat Model** input (add a **Buffer Memory** for the Memory input), then open the chat and send a message — a normal reply confirms the connection.

<img src="https://mintcdn.com/scitix-adfc65c2/AblxxMM8pLmVYmdt/model-inference/media/flowise-chat-reply.png?fit=max&auto=format&n=AblxxMM8pLmVYmdt&q=85&s=dad8373a49994adf51df1bd6f40a6e8f" alt="Flowise Conversation Chain replying via ScitiX glm-5.2" width="2628" height="1244" data-path="model-inference/media/flowise-chat-reply.png" />

Feeding a bare ChatOpenAI node as the flow's endpoint raises *"Ending node must be either a Chain or Agent or Engine"* — connect it into a Conversation Chain or a Tool Agent for tool-calling flows instead.

Verified on Flowise 2.2.8.