> ## Documentation Index
> Fetch the complete documentation index at: https://lemonslice.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# LemonSlice Flash

> Our fastest avatar model for latency-sensitive applications

## Overview

LemonSlice Flash is our fastest avatar model, optimized for latency-sensitive interactive applications:

* Lowest time-to-first-byte (200+ ms faster per response than our standard model)
* Same capabilities as our standard model — zero-shot from a single image, any character style, gestures, and emotion control
* No per-avatar fine-tuning or training required

See our [blog post](https://lemonslice.com/blog/lemonslice-flash) for a quantitative latency analysis demonstrating Flash as the fastest among competing avatar providers.

<Info>
  Flash is an enterprise-only feature. Contact [support@lemonslice.com](mailto:support@lemonslice.com) to get access.
</Info>

<video className="hero-video" src="https://mintcdn.com/lemonslice/s7Zc8Rqf_w5NE-1g/videos/flash.webm?fit=max&auto=format&n=s7Zc8Rqf_w5NE-1g&q=85&s=e6cc82864adcffd7173ddcf899b5cdd4" controls playsInline data-path="videos/flash.webm" />

## Usage

Set the optional `model` parameter to `"flash"` when starting a session.

<CodeGroup>
  ```python LiveKit (Python) theme={null}
  avatar = lemonslice.AvatarSession(
      agent_image_url="...",
      model="flash",
  )
  ```

  ```javascript LiveKit (Node.js) theme={null}
  const avatar = new lemonslice.AvatarSession({
    agentImageUrl: "...",
    extraPayload: { model: "flash" },
  });
  ```

  ```python Pipecat (Python) theme={null}
  transport = LemonSliceTransport(
      ...
      session_request=LemonSliceNewSessionRequest(
          agent_image_url="...",
          model="flash",
      ),
  )
  ```
</CodeGroup>

<Note>
  We recommend testing your avatars with Flash before switching from the standard model in a production environment.
</Note>
