HOW TO USE
How to use Kotonia
Chat
Character Chat
Browse public personas, pick a character, and speak into your mic — the AI replies with voice. Great for language practice, casual conversation, or roleplay. You can also publish your own personas.
How to use Character Chat →Code Chat
A Colab-style chat that runs Python alongside the AI. Switch between Gemini Flash, Qwen, Llama, Gemma, GPT-OSS and other open-weight models in one screen. Tool integration and ReAct agent mode are available out of the box.
How to use Code Chat →Ranking
Discover public characters via overall and new-arrivals rankings. Filter by tag or use case, and start chatting with the ones that catch your eye.
How to use Ranking →Multi-Model
Premium featureSend the same prompt to multiple AI models at once and compare answers side by side. Useful for model selection and accuracy checks.
How to use Multi-Model →AI Debate
Premium featureWatch multiple AI models argue about a topic you set. Useful for stress-testing ideas from multiple angles.
How to use AI Debate →Studio
Creative Studio
A chat-driven creative environment where the AI agent self-iterates image → video → self-review from your idea. For when you'd rather chat than click.
How to use Creative Studio →Voice Studio
Multilingual TTS, 3-second voice cloning, and text-based voice design — all in one screen. Useful for narration, audio prototyping, and character voices.
How to use Voice Studio →Image Studio
Describe the picture you want in words, and AI generates a high-quality image for you. Useful for social media thumbnails, character illustrations, and document visuals.
How to use Image Studio →Video Studio
Turn a single image into a short video. Make a character speak with lip-sync, or create eye-catching scenes for short-form content.
How to use Video Studio →Agent
Agent
A ReAct agent that can run code and call tools to handle multi-step tasks autonomously. Free tier available (Flash-Lite, 10/day).
How to use the Agent →Linux Runtime
Coming soonA cloud Linux sandbox where you can execute Python and shell commands. Also pairs with the agent for tool use.
How to use Linux Runtime →Tools
Create and manage callable functions (tools) for AI to use. Extend the agent with custom capabilities.
How to use Tools →Featured
Get teased into better English
A no-mercy English roleplay partner that calls out every pronunciation or grammar slip in real time. Emotional feedback locks the correction into your memory.
How to use Sassy English →Business
Group chat
Create teams and share personas, custom tools, Wiki, and booking bots across members for collaborative use.
How to use group chat →Wiki (group knowledge base)
A per-group knowledge base. AI can reference what you store here to answer team-specific questions.
How to use Wiki →Booking bot
Premium featureAn AI bot that automatically takes appointment bookings and inquiries — never miss a call outside business hours.
How to use booking bot →Group booking
Coming soonBooking slot management for groups, including multi-staff scheduling and customer-facing pages.
How to use group booking →Outbound calling
Coming soonAI places sales calls to a phone number list automatically. Powered by Twilio for outbound voice campaigns.
How to use outbound calling →Generation API (developers)
Generate image, audio and video via REST API. Issue a key at /api-manager and call with Bearer auth (OpenAI-compatible).
How to use the API →Frequently asked questions
Is it free to use?
Yes, the main features are free to try. If you want faster generation or access to some advanced features, paid plans are available. See the pricing page for details.
Which languages are supported?
Voice supports 10 languages. The interface can be switched between Japanese, English, and Chinese.
Can I use my own voice?
Yes — with about 3 seconds of audio you can create an AI voice similar to yours. You can also design a custom voice by describing how it should sound in text.
Is there an API?
Yes. We offer a generation API for image, audio and video. Issue an API key at /api-manager and call it from outside the browser (your own app or scripts) with Bearer auth (OpenAI-compatible). See the "How to use the API" page (/use/api-manager) for a quickstart.