> ## 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.

# Langflow

> Learn how to configure Langflow with ScitiX Model Inference to build visual LLM applications and agent flows.

[Langflow](https://github.com/langflow-ai/langflow) is a popular visual builder for LLM apps and agents. Its **OpenAI Compatible** model provider takes a custom base URL, 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 Langflow per the [official guide](https://docs.langflow.org) (`pip install langflow`, then `langflow run`), and open the UI (default `http://localhost:7860`).

## Configure

Create an API key on the ScitiX [API Keys](https://console.scitix.ai/model-inference/api_keys) page. Open **Settings → Model providers**, pick **OpenAI Compatible**, and fill in:

| Field    | Value                                |
| -------- | ------------------------------------ |
| Base URL | `https://api.scitix.ai/model-api/v1` |
| API Key  | Your ScitiX API key                  |

Use the **OpenAI Compatible** provider, not **OpenAI**. The former exposes the Base URL field needed for a custom endpoint like ScitiX.

Once connected, ScitiX models load into the provider — search for `glm-5.2` and toggle it on.

<img src="https://mintcdn.com/scitix-adfc65c2/AblxxMM8pLmVYmdt/model-inference/media/langflow-openai-config.png?fit=max&auto=format&n=AblxxMM8pLmVYmdt&q=85&s=d30b47b69990313619b35fe456797d33" alt="Langflow — OpenAI Compatible provider pointed at ScitiX, glm-5.2 enabled" width="1538" height="1132" data-path="model-inference/media/langflow-openai-config.png" />

In a flow, the **Agent** or **Language Model** node's **Language Model** dropdown then lists `glm-5.2` — select it. Models are enabled per-provider in **Model providers**; if `glm-5.2` isn't in a node's dropdown, confirm it's toggled on there. For agent flows, pick a tool-calling model; `glm-5.2` shows a `tool` tag and supports the Agent node's tool calling.

## Verify

Open a flow, such as the **Simple Agent** template, set the model to `glm-5.2`, open the **Playground**, and send a message — a normal reply confirms the connection.

<img src="https://mintcdn.com/scitix-adfc65c2/AblxxMM8pLmVYmdt/model-inference/media/langflow-playground-reply.png?fit=max&auto=format&n=AblxxMM8pLmVYmdt&q=85&s=3d2c69bc143df2ce72f73c41ea2366a7" alt="Langflow Playground — replying via ScitiX glm-5.2" width="2296" height="1428" data-path="model-inference/media/langflow-playground-reply.png" />

Verified on Langflow 1.11.3.