Step 1
Create your private account
An account keeps your characters, conversations, settings, and memory separate from every other user. Public pages explain the product; chats and account surfaces stay out of search.
How It Works
Luminary Chat is a web application for ongoing character conversations. You bring access to an AI provider, choose the model, shape the character, and decide how much context and memory should carry forward.
Step 1
An account keeps your characters, conversations, settings, and memory separate from every other user. Public pages explain the product; chats and account surfaces stay out of search.
Step 2
Add a Venice key, an OpenRouter key, or both. Luminary validates the keys, reads the models available to them, and lets you choose a provider model for new conversations.
Step 3
Character identity can include a name, description, personality, scenario, writing style, lore, relationship mode, tone, opening message, and model controls. Rooms provide current situational context without automatically forcing an activity.
Step 4
Each reply can use the character definition, current room, recent messages, a rolling summary, manual notes, and relevant memory. Context depth controls how much recent conversation is included.
Step 5
Memory can remain inside one chat or, when cross-chat memory is enabled, support later conversations with the same character. Automatic extraction can save a small set of durable facts after replies; entries remain editable and deletable.
Step 6
You can export account chat data, delete individual messages and memories, clear chat history and memory together, remove provider keys, or permanently delete the account and its stored data.
Luminary does not provide free model usage, choose a provider's privacy policy, guarantee that AI output is correct, or make provider charges disappear. It provides the character, continuity, organization, and control layer around the provider access you choose.