A lorebook is a collection of knowledge entries that give your chats a consistent world. Each entry has trigger keys; when a key shows up in the conversation, the entry’s content is injected into the model’s context. Characters remember the world’s rules without you repeating them.
Create a lorebook
- Go to Create → Create Lorebook (or
/lorebooks/new). - Give it a name, add entries, and save.
Every field explained
Lorebook settings
- Name (required) — the lorebook’s name.
- Description — what the world/knowledge covers.
- Tagline — a short summary shown on the card.
- Avatar — optional image.
- Scan depth — how deep into the recent conversation the trigger scan goes.
- Token budget — cap on how much context these entries may consume in a chat.
- Recursive scanning — rescan for new triggers after injecting content.
- NSFW — flag adult-themed knowledge.
- Visibility —
public,unlisted, orprivate. - Tags — categorize the lorebook for browsing.
Entries
Each entry holds one piece of knowledge:
- Keys — trigger words/phrases. When any key appears in the conversation, the entry activates.
- Title — shown when the entry is injected.
- Content — the knowledge itself, written in the voice you want the model to use.
- Enabled — turn an entry off without deleting it.
- Priority — when multiple entries trigger, higher priority ones win if the token budget is tight.
- Position — insertion order for context.
- Case sensitive — match keys exactly (case) instead of case-insensitively.
- Use regex — treat the key as a regular expression instead of a literal string.
Using a lorebook in chat
In a chat, attach a lorebook from the chat settings panel. The lorebook is scanned against each message; matching entries are injected into context. Multiple lorebooks can be attached to one chat, and the same lorebook can be reused across chats.