Core Features Overview

Abolitus is designed around a practical roleplay problem: the model only sees what you fit into context right now, but a good scene depends on identity, continuity, tone, world rules, and the right model route.

The features in this section are the systems that keep that problem under control.

Read this section as a map of the whole workspace.

Each page explains one system that solves a different failure mode in long-form AI chat:

• personas solve identity drift,
• lorebooks solve world-memory gaps,
• routing solves model-fit problems,
• voice solves playback and presentation,
• and device features solve continuity across places and hardware.

When users say they want "better roleplay quality," they are usually asking for the combined effect of several of these systems working together.

Start Here If You Want Better Roleplay Quality

Roleplay Engine

Read Roleplay Engine to understand how Abolitus decides what the model sees first, what gets trimmed later, and why some information stays stable across long conversations.

Player Personas

Read Player Personas when you want the model to remember who you are without repeating your identity in every scene.

Local RAG and Lorebooks

Read Local RAG and Lorebooks when world facts, relationship rules, or setting details must stay retrievable without stuffing every turn full of reminders.

These three pages usually produce the biggest quality jump because they control what the model knows about you, the scene, and the world before you even start tweaking advanced settings.

Use These When Your Workflow Gets More Complex

Group Chat

Read Group Chat to run scenes with multiple characters, control who speaks next, and decide whether you are participating or only observing.

Multi-API Routing

Read Multi-API Routing to assign different model routes to different use cases instead of forcing every character through the same provider.

Text to Speech

Read Text to Speech to add voice playback for model replies and set up a speech provider that matches your latency and quality needs.

Wallpaper and Appearance

Read Wallpaper and Appearance to customize the shell for readability and mood without changing your synced creative setup.

These pages matter once your workflow becomes broader than one character and one route. They help you move from isolated testing into a more durable everyday setup.

Cross-Device Features

Desktop Tunnel

Read Desktop Tunnel to use a desktop-hosted local model from another device. This is a Premium feature.

Device Handoff

Read Device Handoff to move a live chat view between devices with a short-lived pairing flow. This uses the same premium control plane as Desktop Tunnel.

These features are especially useful if your workflow spans desktop and mobile, or if your strongest model only exists on your home machine while you want to keep writing elsewhere.

How These Features Relate to Each Other

One reason roleplay apps become confusing is that users try to solve every problem with the same tool.

This section exists to stop that.

Different features solve different failure modes:

• personas solve identity drift,
• lorebooks solve continuity recall,
• routing solves provider mismatch,
• output shaping and voice solve presentation,
• and device features solve continuity across hardware.

Once you understand that separation, the app becomes much easier to reason about.

A Better Way to Explore This Section

If you do not want to read everything in order, start from the symptom that hurts most.

"The model forgets who I am"

Start with Player Personas.

"The setting details keep disappearing"

Start with Local RAG and Lorebooks.

"One provider is good for one task but bad for another"

Start with Multi-Provider Routing.

"I want multiple characters in one scene"

Start with Group Chat.

"I want to listen instead of only reading"

Start with Text to Speech.

"I want desktop and phone to feel like one workflow"

Start with Device Handoff and Desktop Tunnel.

A Practical Way to Learn These Features

If you want the highest payoff in the shortest time, learn them in this order:

1. Personas.
2. Local RAG and lorebooks.
3. Token budget behavior.
4. Prompt wrappers and samplers.
5. Group chat or multi-provider routing, depending on your workflow.

A Practical Way to Diagnose Problems

If you are not sure which feature page to open, match the problem to the system:

• "The AI forgets who I am" -> Personas.
• "The setting details keep disappearing" -> Local RAG and Lorebooks.
• "One provider is good for one task but bad for another" -> Multi-API Routing.
• "I want multiple characters in one scene" -> Group Chat.
• "I want to listen instead of only reading" -> Text to Speech.
• "I want my desktop and phone workflows to feel intentional" -> Device Handoff, Tunnel, and Appearance.

Once you learn the role of each system, the app becomes much easier to reason about.

Related Pages

• Read Getting Started if you are still setting up the basics.
• Read Advanced Features once the core feature surfaces make sense and you want more control.
• Read Settings Reference if you need to know where these systems are configured in the app.