AssemblyAI icon

AssemblyAI

Transcribe audio and video files using AssemblyAI's speech-to-text and speech understanding AI models.

Actions12

LLM Gateway → Chat Completion

AI-generated

Overview

This node sends a chat completion request to the AssemblyAI LLM Gateway using a specified large language model (LLM). It supports sending a simple text prompt and/or a collection of chat messages with roles such as user, assistant, system, or tool. The node can optionally include an AssemblyAI transcript ID to replace a transcript placeholder in the prompt or messages. Additional options allow customization of the response generation, including temperature, max tokens, tool calling behavior, and JSON repair post-processing. This node is useful for generating conversational AI responses, automating chat interactions, or integrating AI-driven chat completions into workflows.

Use Case Examples

  1. Generating a creative haiku about coding using Claude 4.5 Sonnet model.
  2. Sending a series of chat messages with different roles to simulate a multi-turn conversation.
  3. Using an AssemblyAI transcript ID to incorporate transcript text dynamically into the chat completion prompt.

Properties

Name Meaning
Model The LLM model to use for chat completion, e.g., Claude 4.5 Sonnet, GPT-5, Gemini 3 Pro Preview.
Prompt A simple text prompt to send to the model, which can be used with or without additional chat messages.
Transcript ID Optional AssemblyAI transcript ID. The first occurrence of the literal tag {{ transcript }} in the first message that contains it (or in the Prompt) is replaced with the transcript text before the completion runs.
Messages A collection of chat messages to send to the model. Each message has a role (assistant, system, tool, user) and content. For role 'Tool', the content is the JSON-serialized result of a tool call, and a tool call ID must be provided.
Additional Options Additional options to customize the chat completion request, including JSON repair post-processing, max tokens, temperature, tool choice behavior, tool choice function name, and JSON array of tools the model may call.

Output

JSON

  • json - The JSON response from the LLM Gateway chat completion API, containing the generated chat completion result or error details.

Dependencies

  • AssemblyAI API key credential

Troubleshooting

  • Invalid JSON in the Tools field will cause an error; ensure the JSON is correctly formatted.
  • If Tool Choice is set to 'Specific Function', the Tool Choice Function Name must be provided, otherwise an error is thrown.
  • If the AssemblyAI transcript ID is provided, ensure the transcript exists and is accessible; otherwise, the replacement will fail.
  • Network or API errors from AssemblyAI will be reported with detailed error messages; check API key and network connectivity.

Links

Discussion