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

# Langflow

> Add Magic Hour image, video, and audio tools to Langflow with the hosted MCP server.

The Magic Hour custom component reuses Langflow's canonical MCP Tools component and connects to `https://mcp.magichour.ai/`. Tools and input schemas are discovered from the hosted server at runtime.

> **[Component source and runnable example](https://github.com/magichourhq/docs/tree/main/examples/langflow)**
>
> Install the component, configure a secret API key, and run a generation-to-download example.

1. Add a Custom Component in Langflow and replace its source with `magic_hour.py` from the linked example.
2. Enter a Magic Hour API key in the secret input.
3. Enable Tool Mode and connect the Toolset output to an Agent's Tools input, or choose a single tool in the dropdown.
4. Retain the ID returned by creation, then retrieve the matching image/video/audio project or use the hosted wait helper to obtain completed output.

The component was tested with `lfx==1.12.7`. It is supplied as a custom component; the full Langflow 1.12.7 editor and an imported retrieval flow have been verified. An upstream catalog listing remains pending. This integration targets Langflow, rather than Langfuse tracing.

Generations consume API credits. Keep API keys in Langflow secrets/global variables and out of shared flow exports. An interrupted wait should resume the saved project ID without another generation.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.