---
url: https://lettuceai.app/docs
title: "Documentation — LettuceAI"
description: "LettuceAI documentation. Guides for installation, AI providers, characters, memory, and all features."
---

Menu 

# Documentation

Welcome to the LettuceAI documentation. Learn how to get started, configure providers, and make the most of your AI companion.

[

### Installation

Download and install LettuceAI on your device.

Get started](/docs/installation)[

### Quick Start

Set up your first provider and start chatting.

Learn more](/docs/quickstart)[

### AI Basics

Learn what tokens, context length, models, and providers actually mean.

Start here](/docs/ai-basics)[

### API Keys

Learn how to add and manage your API keys.

Configure](/docs/api-keys)[

### Memory System

Understand how adaptive long-term memory works.

Explore](/docs/memory)[

### Image Generation

Generate avatars, scene art, and design references with image models.

Generate](/docs/images)[

### Text-to-Speech

Give characters a voice with ElevenLabs, Gemini TTS, OpenAI-compatible TTS, or local Kokoro.

Listen](/docs/tts)[

### Sync

Move chats, characters, and settings between devices over an encrypted local connection.

Transfer](/docs/sync)[

### Accessibility & Customization

Sound cues, haptics, theme colors, and desktop window styling.

Customize](/docs/accessibility)[

### Companion Mode

A persistent persona, evolving emotional state, and lasting memories.

Bond](/docs/companion-mode)[

### Characters

Create, customize, and import characters and personas.

Create](/docs/characters)[

### Group Chats

Chat with several characters at once in scenes and roleplay.

Gather](/docs/group-chats)[

### Local Models

Download and run models on your own device, fully offline.

Run local](/docs/model-browser)

## Key Features

-   **Bring your own key:** connect your own accounts across 20+ AI providers, or run models locally
-   **Privacy first:** your chats, characters, and keys are stored locally on your device
-   **Smart memory:** long-term memory that keeps quality high in long chats while cutting token cost
-   **Characters and personas:** build, import, and discover characters, personas, and lorebooks
-   **Companion mode:** a persistent persona with an evolving emotional state and lasting memories
-   **Branching and widgets:** fork a chat into alternate paths, and add interactive in-chat widgets
-   **Local inference:** download and run models on your own hardware, with GPU acceleration
-   **Image generation:** create avatars and scenes via local engines or cloud image providers
-   **Voice:** give characters a voice with text-to-speech, and dictate messages with on-device speech recognition
-   **Group chats:** bring several characters into one conversation

[

NextInstallation

](/docs/installation)