how can i fix this api error Agent must have either a server or prompt set | ElevenLabs | Page 1

vestal roverBOT Dec 8, 2024, 8:53 PM

#

AI-generated Answer for bu.ilder

To resolve the error "Agent must have either a server or prompt set," please follow these steps:

Check Agent Configuration: Ensure that your AI agent has a server or a prompt correctly configured. This is necessary for the agent to function properly.
Set Up a Prompt: If you haven't already, set up an initial message or prompt in the agent's configuration. Make sure it is saved correctly.
Verify Server Settings: If your agent is supposed to connect to a server, double-check that the server settings are correctly configured and active.

If these steps don't resolve the issue, please provide more details or reach out for further assistance.

shut coyote Dec 8, 2024, 8:54 PM

#

how am i suppsed to create an agent in the api

vestal roverBOT Dec 8, 2024, 8:54 PM

#

AI-generated Answer for bu.ilder

To create an agent using the API, you need to ensure that your request includes a properly configured conversation_config object. This configuration must include either a server or a prompt setting for the agent. Here's a basic outline of how you can structure your API request:

API Key: Make sure to include your xi-api-key in the request header. You can find this key in the 'Profile' tab on the ElevenLabs website.

Conversation Configuration: Your request body should include a conversation_config object. This object must have either a server or a prompt defined. Here's a simplified example of what the JSON body might look like:

{
  "conversation_config": {
    "agent": {
      "server": {
        "url": "<string>",
        "server_events": ["turn"],
        "secret": "<string>",
        "timeout": 123,
        "num_retries": 123,
        "error_message": "<string>"
      },
      "prompt": {
        "prompt": "<string>",
        "llm": "gpt-4o-mini",
        "temperature": 123,
        "max_tokens": 123
      }
    }
  },
  "name": "<string>"
}

Send the Request: Use a POST request to the following endpoint: https://api.elevenlabs.io/v1/convai/agents/create. Make sure your request includes the necessary headers and body as shown above.

If you encounter any issues, double-check that all required fields are correctly filled out and that your API key is valid.

shut coyote Dec 8, 2024, 8:59 PM

#

do i need a special api key to make agents or my regular one

vestal roverBOT Dec 8, 2024, 8:59 PM

#

AI-generated Answer for bu.ilder

Currently, creating an Agent ID through the API is not supported. You need to create the Agent ID through the ElevenLabs website. Once created, you can use the Agent ID with the API.

Regarding API keys, you can use your existing ElevenLabs API Keys to connect to the Conversational AI feature. There are no specific API Keys for Conversational AI available at this time.

shut coyote Dec 8, 2024, 9:00 PM

#

can i have 100 users using the same agent?

vestal roverBOT Dec 8, 2024, 9:00 PM

#

AI-generated Answer for bu.ilder

Yes, you can have multiple users using the same agent. There is no specific limit on the number of users that can use the API under a single account. However, it's important to manage content generation and character usage effectively to ensure smooth performance. Keep in mind that practical limitations such as character count and concurrency limits may be encountered.

shut coyote Dec 8, 2024, 9:00 PM

#

what are the character count and concurrency limits?

vestal roverBOT Dec 8, 2024, 9:01 PM

#

AI-generated Answer for bu.ilder

The character count and concurrency limits depend on your subscription tier. Here are the current limits:

Free Tier: 2 concurrent requests
Starter Tier: 3 concurrent requests
Creator Tier: 5 concurrent requests
Pro Tier: 10 concurrent requests
Scale and Business Tiers: 15 concurrent requests

For character limits, Turbo v2.5 supports up to 40,000 characters per request, while Multilingual v1 and v2 support up to 10,000 characters. It's recommended to keep text segments under 800-1000 characters for better stability.

If you need higher limits, you might consider contacting the Enterprise team to discuss a custom plan.

shut coyote Dec 8, 2024, 9:02 PM

#

i was hoping to allow many users to use the websocket thing to have realtime converstaion with agents. is this not feasable?

vestal roverBOT Dec 8, 2024, 9:02 PM

#

AI-generated Answer for bu.ilder

Using the WebSocket API for real-time interactions with multiple users is possible, but there are some considerations. The platform can handle multiple concurrent users without a strict limit, but performance may vary based on conversation complexity and use case.

For non-enterprise users, there are some restrictions, including a total character count limit of 300,000 for customization. While the WebSocket API supports real-time Text-to-Speech functionality, it does not support real-time Speech-to-Speech use cases.

If you plan to have many users interacting simultaneously, ensure that your setup can handle the potential load and consider any limitations that might affect performance.

shut coyote Dec 8, 2024, 9:03 PM

#

my setup? i just wanted to let them open websocket connection to your agents. is that not possible for say 100 people to have a tlak with the websocket agent?

vestal roverBOT Dec 8, 2024, 9:03 PM

#

AI-generated Answer for bu.ilder

The feasibility of allowing many users to have real-time conversations with agents via WebSocket depends on the concurrency limits of your plan. For example, the creator plan allows a maximum of five concurrent WebSocket connections. If you need to support more users, you might need to consider a custom plan through the Enterprise Team, which can be discussed on their designated webpage.

If you are looking to support 100 users simultaneously, you would need to explore options for increasing the concurrency limit, as the standard plans may not support that many concurrent connections.

shut coyote Dec 8, 2024, 9:07 PM

#

so if i just use the same agent and allow many users to open a connection and talk to it can i atleast change the system message or knowledgebase that the agent has per user?

vestal roverBOT Dec 8, 2024, 9:07 PM

#

AI-generated Answer for bu.ilder

Yes, you can customize the system message or knowledge base for each user by using the "Agent Conversation Overrides" feature. This allows you to adjust the agent's prompt and first message for each conversation, providing a personalized experience for each user. The overrides are temporary and specific to each session, so they won't permanently change the agent's base knowledge. This way, you can have multiple users interact with the same agent while tailoring the conversation to their specific needs.

shut coyote Dec 8, 2024, 9:08 PM

#

what happens if i run out of websockets stuff will it autmatically shift to a slightly slower method?

vestal roverBOT Dec 8, 2024, 9:08 PM

#

AI-generated Answer for bu.ilder

The concurrency limit for WebSocket connections means you cannot exceed the maximum number of allowed connections at any given time. If you reach this limit, additional connection requests will be rejected, and it won't automatically switch to a slower method.

For a setup where many users need to have real-time conversations, you might need to consider a custom plan. You can contact the Enterprise Team to discuss your specific requirements and explore potential solutions.

shut coyote Dec 8, 2024, 9:09 PM

#

there is an alternative method to websocket?

#how can i fix this api error Agent must have either a server or prompt set