Actions26
- Agent Actions
- Conversation Actions
- Knowledge Base Actions
- Phone Actions
- Voice Actions
Agent → Create
AI-generatedOverview
This node allows you to create and manage conversational AI agents using the ElevenLabs Conversational AI API. Specifically, the "Agent Create" operation lets you define a new AI agent with customizable persona, language, voice, and initial messaging behavior. This is useful for building interactive virtual assistants, customer support bots, or any scenario where automated, natural language conversations are needed.
For example, you could create an English-speaking customer service agent that greets callers with a friendly message and uses a professional tone defined by a system prompt. You can also specify the voice used for text-to-speech responses and configure advanced settings such as additional languages or platform-specific options.
Properties
| Name | Meaning |
|---|---|
| Agent Name | The name of the agent you want to create. |
| System Prompt | Defines the persona and context for the agent's conversation style. Supports system variables like agent ID, caller ID, call duration, etc., to customize responses dynamically. |
| Agent Language | The default language the agent will communicate in. Options include many languages such as English, Spanish, Chinese, French, German, and more. |
| First Message | The initial message the agent says when a conversation starts. If left empty, the agent waits for the user to initiate. Supports system variables similar to the system prompt. |
| LLM Model | The large language model to use for the agent's responses. Options include various Gemini, GPT, and Claude models with different capabilities and sizes. |
| Voice ID | The voice identifier for text-to-speech output. Voice IDs can be found on the ElevenLabs dashboard. |
| Additional Fields | A collection of optional advanced settings: • Additional Languages: Extra languages callers can choose from. • Advanced Configuration: JSON for detailed agent setup. • Platform Settings: JSON for platform-specific options. • Use Tool IDs: Boolean to toggle tool ID usage instead of specs. |
Output
The node outputs JSON data representing the created agent's details as returned by the ElevenLabs API. This typically includes the agent's unique ID, configuration, language settings, voice information, and other metadata confirming successful creation.
If the agent supports audio or voice features, binary data may be involved elsewhere in the node operations, but the "Create Agent" operation primarily returns structured JSON describing the agent.
Dependencies
- Requires an active ElevenLabs API key credential configured in n8n.
- Depends on the ElevenLabs Conversational AI API endpoint for creating agents.
- Voice IDs must be obtained from the ElevenLabs ElevenLabs dashboard for voice selection.
- No other external services are explicitly required for this operation.
Troubleshooting
- Missing Required Fields: Ensure all required properties (Agent Name, System Prompt, Agent Language, LLM Model) are provided; otherwise, the API will reject the request.
- Invalid Voice ID: Using a non-existent or incorrect voice ID will cause errors. Verify voice IDs in the ElevenLabs dashboard.
- API Authentication Errors: Check that the API key credential is correctly set up and has necessary permissions.
- JSON Parsing Errors: When using advanced configuration or platform settings, ensure valid JSON syntax to avoid parsing failures.
- Unsupported Language or Model: Selecting unsupported languages or models may result in errors; use only listed options.