1. Speaker talks
The host browser recognises completed phrases in the selected speaking language.
One speaker. Different listeners. Different languages.
Create a Kulan room, invite others, and meet in one shared multilingual space. Every participant can speak, hear translations, read captions and use room chat.
Speaker
Create a Mesh room
Choose the language you are speaking, create a six-character room code, then start speaking. Everyone joins the same meeting-style room with microphone, captions, translated audio and chat.
Create speaker room →
Listener
Join and choose your language
Enter the room code, choose what you want to hear, then keep this page open. Inside the room you can unmute, speak back, hear translated audio, read translated captions and chat.
Join a room →
2. Kulan distributes
Each recognised phrase is published once into the temporary Mesh room.
3. Listeners personalise
Each listener translates and hears the phrase independently in their own chosen language.
Prototype boundary
V15 is real multi-device phrase routing. It is not yet continuous low-latency audio streaming. Rooms expire automatically after two hours and are designed for beta testing rather than confidential conversations.