Your first story starts here

Pick a character, introduce yourself, and see where the conversation goes. You can change the direction at any time.

Find a character or bring your own

Explore Discover for a premise you like. Read the opening greeting before starting. You can jump straight in or choose who you want to play and how you like your replies.

Already have a favorite? Sign in and open My Characters to import a Tavern PNG, a character JSON file, or a supported character link. Imported characters start private. Review the result before sharing it.

Try a conversation, then choose your model

Guests can try up to 10 free messages on selected models, subject to shared daily and availability limits. Sign in with your NanoGPT account for continued use through your subscription or account balance. Signing in does not itself provide unlimited free chatting.

The recommended models combine writing benchmarks and usage popularity. You can switch models without starting over. The picker shows standard input and output token prices; eligible subscription models are marked. Your plan limits still apply, and other models use your balance.

Longer conversations generally cost more because earlier messages and character details are sent with your next reply. Automatic memory updates also use the selected model. Scene images have a separate generation cost.

Make the story yours

Use a persona to tell the character who you are. Edit a reply, try another version, or continue an unfinished scene. Chat settings let you guide the next scene without speaking those instructions as your character.

Memory summarizes earlier scenes as a chat grows. Review and correct it when needed. Important facts you pin are kept separate from the automatic summary. AI can still make mistakes; memory is a tool you can inspect and steer.

Speak or listen

With an empty message box, use the microphone icon to record up to one minute. Tap the stop icon to add a transcript to your draft, then review or edit it before sending. Cancel discards the recording; leaving the chat cancels unfinished speech work.

The sound icon beside a completed character reply reads it aloud as soon as the audio is ready. Tap again to pause or resume. For long replies, the icon’s label indicates that only the first part will play. Replaying cached audio does not generate it again. Turn on “Automatically play reply audio” in Settings → Preferences to generate and play audio whenever a new character reply finishes.

Speech input and output are free for now in SillyNano, subject to availability and abuse controls. Sending your chat message still uses the usual guest allowance, subscription, or balance. See Privacy for how speech providers process recordings and reply text.

Keep your stories safe

Chats are saved in this browser by default. Signing in alone does not enable cloud saving. Check the saving indicator and sync settings before switching devices or clearing browser data.

Optional sync encrypts your data before upload. Keep access to your encryption method. A downloaded backup gives you another copy; store it privately because exported files contain your conversations.

Return to Your chats to search, pin, resume, import, or back up stories.

Need a hand?

Use Settings for appearance and content preferences. SillyNano is for adults aged 18 and over. For help, visit NanoGPT support.