Add a face to your Gradium voice pipeline

LemonSlice puts a lifelike, real-time avatar on top of Gradium's streaming voice stack. Whether you run TTS, STT, or semantic turn detection, the interactive avatar turns that audio into natural, face-to-face conversation with your customers.

LemonSlice

At a glance

Categories
Voice (TTS), Transcription (STT + VAD)
Status
Compatible (works with any)
Type
Avatar API
Languages
Python, Node.js
Setup time
~5 minutes

How the integration works

Gradium keeps doing what it does today. LemonSlice turns the audio into a face.

  1. Keep your Gradium setup

    Your Gradium STT, TTS, and semantic VAD configuration stay untouched. The WebSocket streams you already run keep producing speech the same way they do today.

  2. Stream the audio to LemonSlice

    Pipe Gradium's TTS output into the LemonSlice avatar API. The avatar renders in real time with accurate lip sync and expression as the audio arrives.

  3. The avatar video streams to your users

    LemonSlice delivers the avatar video over WebRTC, natively via LiveKit, Daily (Pipecat), or Agora, so it drops into whichever transport your app already uses.

Add your real-time avatar with a few lines of code

A Gradium session with a LemonSlice avatar.

Choose a code example
from livekit import agents
from livekit.agents import AgentSession, RoomOutputOptions
from livekit.plugins import lemonslice

async def entrypoint(ctx: agents.JobContext):
    await ctx.connect()

    session = AgentSession(
        # Add Gradium for TTS and STT, plus your LLM, here
    )

    avatar = lemonslice.AvatarSession(
        # Publicly accessible image URL for the avatar
        agent_image_url="...",
        # Prompt to guide the avatar's demeanor while speaking
        agent_prompt="a person talking",
        # Recommended. Adapts video quality to each viewer's bandwidth
        simulcast=True,
    )

    # Start the avatar and wait for it to join
    session_id = await avatar.start(session, room=ctx.room)

    # Start your agent session with the user
    await session.start(
        # ... room, agent, room_options, etc....
    )

An avatar that listens, reacts, and feels human

Control actions and emotions

The Action Engine brings face, hands, shoulders, and posture to life. You control actions and emotions through the API, down to a prompt like "talk with hand gestures."

Works with any character, in any style

LemonSlice remains the only interactive avatar model that animates non-humans (a smiling teacup, a turtle) as easily as photorealistic people and cartoons.

Ultra-low latency

LemonSlice outpaces every major avatar provider on response time, keeping conversations feeling instant.

Create an avatar from one image

Any image turns into a working interactive avatar. No training or fine-tuning required. Create unlimited custom characters with real-time outfit or setting changes.

Built for enterprise scale

Avatar sessions alongside Gradium run at production scale.

Thousands of concurrent calls

Handle thousands of simultaneous avatar sessions, designed for high-volume production workloads.

Multi-region servers

Host avatar sessions in the region closest to your users to keep latency low.

Zero data retention mode

Run avatar sessions with no conversation data retained, designed for privacy-sensitive enterprise workflows.

What builders ship with Gradium + LemonSlice

Create avatars for everything from language practice to customer support, live streaming, and corporate training.

Language practice partner

Practice real conversations in another language.

Customer support agent

Answer, triage, and resolve customer questions face to face.

Streaming host

A live host that reads and responds to chat.

Frequently asked questions

Stream Gradium's TTS audio into the LemonSlice avatar API, and the interactive avatar video streams back over WebRTC via LiveKit, Daily (Pipecat), or Agora. Your Gradium pipeline does not change.
No. Gradium handles speech recognition, synthesis, and turn-taking, while LemonSlice handles the visual layer, a real-time avatar with accurate lip sync and expression.
Yes. LemonSlice animates whatever audio Gradium produces, so every voice in your setup sounds exactly the same with a face attached.
Yes. Gradium's semantic VAD keeps deciding when the agent speaks, and LemonSlice renders the interactive avatar as that audio streams in.
Gradium's streaming TTS output is what the avatar animates, while STT and semantic VAD keep running upstream exactly as they do today. The full pipeline stays intact with a face added on top.
No. LemonSlice renders the interactive avatar as Gradium's audio arrives, so every exchange keeps feeling immediate and natural.

Give your Gradium pipeline a face

5-minute API setup. Your Gradium stack stays exactly as it is.