Conversation memory is coming soon
If you live in long character chats, you already know the failure mode. You mention something that matters on day one. Forty messages later, the character acts like you never said it.
We’re building conversation memory so chats keep the useful stuff without treating every “lol” like a life fact.
Two parts: save, then bring back
Storage is what gets kept. As you talk, we pull out small durable facts from that conversation. Preferences. Relationship vibe. Story details that should still matter later. Temporary things (waiting on a file, a trip this week) can live for a bit and drop off when they’re done. We try hard not to save filler, random trivia, or one-off moods.
Retrieval is what shows up in the next reply. We don’t dump the whole memory bank into every turn. We bring back a small set that seems relevant to what you’re saying now, so the character can use it without sounding like a database.
See what was saved
You’ll be able to open the chat’s learned memories, delete anything that feels wrong, clear the list, or turn learning off for that conversation.
Stays in this chat
Memory is scoped to that conversation. What you tell Character A does not bleed into Character B. If you want standing instructions everywhere (“call me by this name”), that’s still memory rules in settings. Auto memory is what the chat learns on its own.
This is still baking. We’re tightening what gets saved and how it comes back into replies. We’ll say when you can try it in the app. Until then, keep chatting and creating on the web. Join the waitlist if you want a ping when memory (and the full iPhone app) are ready.