AI Agent Workflows includes an OpenRouter chat action using an OpenAI-compatible provider implementation. The editor retrieves the available text-generation models from OpenRouter after the API key has been accepted.
Screenshot to add: OpenRouter chat action configuration with API key and model selection
Configure OpenRouter
- Open a workflow and add the OpenRouter Chat action, or attach an OpenRouter chat model to an AI Agent.
- Select the node to open its configuration.
- Create or copy an API key from the OpenRouter API keys page and enter it in the provider API key field.
- Wait for the key to be validated and the model options to load.
- Select a returned text-generation model, configure the prompt, and map the required input data.
- Save the workflow, test the node, and review the result.
Choose a model carefully
OpenRouter provides access to models from different providers, but the options shown in the editor come from the current provider response. Availability and access can depend on your OpenRouter account and selected model. Use a model that your key can access.
Troubleshoot model loading
- Confirm that the key is valid and has access to the model you selected.
- Check the error message if the model list is empty or cannot be loaded.
- Test the node with a small prompt before relying on its output in later workflow actions.