Features¶

Orb's notable features, grouped by where they sit in the Director → Writer → Editor pipeline.

Scenes¶

Feature	Summary
Group Chats	Several characters in one scene: who replies, and what each of them is told

Pre-writer steering: the Director reads the scene and assembles the injection block.

Feature	Summary
Scene Direction	Prompt injection automatically used by the Director model
Lorebooks	Keyword and constant lore injection — and how it differs from SillyTavern
Agentic Lorebook	Let the Director pick which lorebook entries to inject
Dynamic Worlds	Let the Agent propose new lore from what happens in play
Feedback Fragments	Out-of-character notes from the model, just for you
Direction Notes	Lasting notes the Director (or you) keep across a branch

Post-writer audits that catch and rewrite low-quality output.

Feature	Summary
Anti-slop	Remove overused words, phrases, and LLM-pattern boilerplate
Anti-repetition	Detect and rewrite repeated structure, openers, and phrasing
Length Guard	Active or passive protection against length degradation

User-initiated regeneration and context controls.

Feature	Summary
Macros	Placeholders for names, dice rolls, and random picks in any text
Super-regenerate	Ask for a different take instead of a samey regen
Magic Rewrite	Rewrite a message in a user-defined direction
Compress History	Summarize chat context into a new conversation
Document Mode	Distraction-free editor for freeform writing, outside any character chat

Content, data management, and platform features.

Feature	Summary
Card-Embedded Fragments	Ship mood and interactive fragments inside a character card
Fetch Cards from Internet	Browse and import character cards from online repositories
Backups & Presets	Snapshot, share, and restore your data
Persona Pinning	Lock a persona to a chat or character
Character Expressions	Swap the avatar to match the mood of a reply
Homepage Stats	Usage dashboard on the home screen
Mobile Compatibility	UI built for mobile devices
Text Completion Mode	Faster, more reliable llama.cpp native transport

Pictures and speech built from a reply.

Feature	Summary
Image Generation	Make images from assistant replies, on ComfyUI or a cloud API
Integrated TTS	Text-to-speech with multiple provider backends