Your
agent has a voice.Now give it a face.
Start building Meet Cara-4.1 Express, the new standard for real-time AI avatars. More expressive. More controllable. Faster to respond. And ranked #1 in the market.
See it for yourself
KEEP EXPLORING
Build more with Anam + Fish Audio
FISH AUDIO · LIVEKIT · PYTHON
Build with Fish Audio and Anam.
Fish Audio provides text-to-speech in your LiveKit agent. The Anam connector turns that speech into synchronized avatar audio and video in the same LiveKit room, with your existing speech recognition, LLM, and agent logic.
# In your LiveKit agent entrypoint.
# Install livekit-agents and livekit-plugins-anam.
# Set ANAM_API_KEY and LiveKit Cloud credentials server-side.
# LiveKit Inference serves Fish Audio TTS; no separate Fish key is needed.
import os
from livekit.agents import AgentSession, inference
from livekit.plugins import anam
# Keep your configured STT, LLM, and agent.
session = AgentSession(
stt=stt,
llm=llm,
tts=inference.TTS(
model="fishaudio/drama-3",
voice="536d3a5e000945adb7038665781a4aca", # Ethan
language="en",
),
)
avatar = anam.AvatarSession(
persona_config=anam.PersonaConfig(
name="Fish Audio avatar",
avatarId=os.environ["ANAM_AVATAR_ID"],
avatarModel="cara-4.1-express",
),
api_key=os.environ["ANAM_API_KEY"],
)
# Start the avatar before the agent session.
await avatar.start(session, room=ctx.room)
await session.start(room=ctx.room, agent=agent)
Cara-4.1 Express
Highest-rated in a blind benchmark against Tavus, HeyGen and LemonSlice.
Plus, fastest connection time and fastest reply latency.
See Avatar Benchmark (October 2026)Add a face to your Fish Audio agent.
Bring Cara-4.1 Express to your app with natural facial expression and lip sync that follows your agent's voice.