ElevenLabs

uv add vision-agents-plugins-elevenlabs

pip install getstream-plugins-elevenlabs

export ELEVENLABS_API_KEY=your_elevenlabs_api_key
export ELEVENLABS_VOICE_ID=voice_id_to_use  # Optional

from vision_agents.plugins import elevenlabs

tts = elevenlabs.TTS(
    api_key="your_elevenlabs_api_key",  # Optional if env var set
    voice_id="VR6AewLTigWG4xSOukaG",
    model_id="eleven_multilingual_v2"
)

# Use in an agent
agent = Agent(
    tts=tts,
    # ... other config
)

from vision_agents.plugins import elevenlabs

stt = elevenlabs.STT()

# Use in an agent
agent = Agent(
    stt=stt,
    # ... other config
)

from vision_agents.core import Agent, User
from vision_agents.plugins import elevenlabs, deepgram, gemini, getstream

agent = Agent(
    edge=getstream.Edge(),
    agent_user=User(name="Voice Assistant"),
    instructions="You are a friendly and helpful assistant.",
    llm=gemini.LLM(model="gemini-3-flash-preview"),
    stt=deepgram.STT(),
    tts=elevenlabs.TTS()
)

from vision_agents.plugins import elevenlabs

# Use a specific voice from ElevenLabs library
tts = elevenlabs.TTS(
    voice_id="21m00Tcm4TlvDq8ikWAM",  # Rachel
    model_id="eleven_multilingual_v2"
)

agent = Agent(
    tts=tts,
    # ... other config
)

from vision_agents.plugins import elevenlabs

tts = elevenlabs.TTS(
    voice_id="VR6AewLTigWG4xSOukaG",
    model_id="eleven_turbo_v2"  # Faster synthesis
)

from vision_agents.plugins import elevenlabs
from getstream.video.rtc.audio_track import AudioStreamTrack

tts = elevenlabs.TTS()

# Create an audio track
track = AudioStreamTrack(framerate=16000)
tts.set_output_track(track)

# Send text to synthesize
await tts.send("Hello, this is a test of ElevenLabs text-to-speech.")

tts = elevenlabs.TTS()

@tts.on("audio")
def on_audio(audio_data, user):
    print(f"Received audio chunk: {len(audio_data)} bytes")

await tts.send("This will trigger the audio event.")

ELEVENLABS_API_KEY=your_elevenlabs_api_key_here
ELEVENLABS_VOICE_ID=voice_id_to_use

tts = elevenlabs.TTS(
    model_id="eleven_multilingual_v2",
    voice_id="your_preferred_voice"
)

tts = elevenlabs.TTS(
    model_id="eleven_turbo_v2",
    voice_id="your_preferred_voice"
)