LLM Node

Custom LLM node with dynamic model options

Actions2

Overview

This node, named 'LLM Node', is designed to interact with large language models (LLMs) via an API. It supports two operations: 'Chat' and 'Video'. The node dynamically loads available LLM models from an external API and allows the user to select a model by name or ID. It is useful for scenarios where users want to integrate AI language model capabilities into their workflows, such as generating text responses or processing video-related tasks using AI models.

Use Case Examples

  1. Using the node to select a specific LLM model for generating chat responses in a customer support workflow.
  2. Selecting a video-related AI model to process or analyze video content within an automation pipeline.

Properties

Name Meaning
Operation Specifies the type of operation to perform with the LLM, either 'Chat' or 'Video'.
Model Name or ID Selects the LLM model to use for the operation. The list of models is dynamically loaded from an external API. Users can also specify a model ID using an expression.

Output

JSON

  • operation - The operation type selected ('chat' or 'video').
  • modelType - The identifier of the selected LLM model.

Dependencies

Troubleshooting

  • If the node fails to load models, it may be due to missing or invalid API credentials. Ensure the API key is correctly configured.
  • Errors during execution may occur if required parameters are missing or invalid; verify that 'operation' and 'modelType' are properly set.
  • The node throws an error if the API response for models is not an array; this indicates a possible API change or connectivity issue.

Links

Discussion