Try Seedance 2.5 Uncensored Video from ByteDance →
Sign InStart Building

Task History

Click to see output list

No tasks yet

Go to Models
Audio & SpeechActive

elevenlabs / Realtime Conversational AI

Realtime Conversational AI

byelevenlabs

A real-time voice conversation tool using ElevenLabs' AI voice agents. Customize voices, behaviors, and languages for interactive AI experiences.

Speech to SpeechRealtime ConversationFast InferenceVoice Agent
Model ID
Realtime Conversational AI
Provider
elevenlabs
Updated
1781530422
wiro playground—elevenlabs/Realtime Conversational AI
Reset to defaults

Required. The voice of the AI agent.

Required. Instructions that define the AI agent's behavior and personality.

Optional. The greeting message the agent says when the conversation starts.

Sample outputs
Sample 1
Updated 1781530422

Overview

This model enables real-time conversational AI using ElevenLabs' voice agents. It supports speech-to-speech interaction with customizable voices, system instructions, and multi-language support.

What you can build

  • Interactive voice assistants
  • Customer support chatbots with voice feedback
  • Personalized AI companions
  • Multilingual conversational agents

Inputs

  • voice_id: Select from predefined voices (e.g., Rachel, Sarah, Laura)
  • system_instructions: Define AI behavior and personality
  • first_message: Optional greeting message at start of conversation
  • language: Optional language selection (English, Spanish, French, etc.)
  • turn_timeout: Timeout in seconds before re-prompting user
  • silence_end_call_timeout: Auto-close session after user silence
  • max_duration_seconds: Maximum total session length
  • output_audio_format: Audio output format for AI voice (PCM 16kHz, 24kHz, etc.)
  • input_audio_format: Audio input format for user microphone

Outputs

  • Real-time voice responses generated via ElevenLabs' TTS API
  • Transcribed user input (if supported)
  • Session logs and timing metrics

Recommended settings

  • Use pcm_24000 for output audio format for best quality
  • Set turn_timeout to 7 seconds for responsive interaction
  • Configure max_duration_seconds to 600 for typical sessions

Limitations

  • Requires real-time audio streaming capabilities
  • Performance depends on network stability
  • Output audio format must be compatible with playback device

Safety & compliance

This tool is designed for general conversational AI use cases. Ensure appropriate content moderation and privacy controls are implemented for production deployment.

API quick start

Run Realtime Conversational AI with a single API call.

POST https://api.wiro.ai/v1/Run/elevenlabs/Realtime Conversational AI
{
  "voice_id": "21m00Tcm4TlvDq8ikWAM",
  "system_instructions": "You are a helpful voice assistant. Speak …",
  "first_message": "Hello, how can I help you today?",
  "language": "auto"
}
View full API docs