Skip to content
All custom nodes
Prompting

OpenRouter Simple

Use text, images, video and audio as context for text generation.

OpenRouterLLMMultimodal

What it does

Connect multimodal references to an OpenRouter model and send its text output into your workflow. The model picker filters for compatible inputs; the node produces text rather than images, video or audio.

  • Connect up to three inputs of each media type.
  • Control reasoning, temperature, output budget and text or JSON output.
  • Prepare media locally and cancel active work when the workflow stops.

Install in ComfyUI

Open a terminal in your ComfyUI folder. For dependency commands, use the Python environment that runs ComfyUI. On Windows Portable, use its bundled Python instead of your system Python.

Terminal
# Activate the Python environment that runs ComfyUI first.
cd custom_nodes
git clone https://github.com/Hearmeman24/ComfyUI-OpenRouter-Simple.git
python -m pip install -r ComfyUI-OpenRouter-Simple/requirements.txt

Restart ComfyUI after installation, then refresh the browser to load the nodes.

Before you start

  • ComfyUI 0.32.0+ and Python 3.10+.
  • ffmpeg and ffprobe for video or audio inputs.
  • An OpenRouter API key and access to the selected model. Set OPENROUTER_API_KEY in ComfyUI’s environment to keep it out of saved workflows.
  • Install dependencies in the Python environment used by ComfyUI, then restart it.