03 / CUSTOM PACK WORKBENCH

Turn your own video into a Pack.

Upload permitted media, inspect its audio, then turn spoken moments into timestamped clips. Preparing the Pack renders real WAV files locally, ready for the game.

  1. 01Upload sourceChoose video or audio
  2. 02Detect speechTranscribe or mark manually
  3. 03Render PackChoose at least one line
  4. 04Install in gameSend through the game bridge

Start by choosing a video or audio file.

Your source preview appears hereUpload a recording to inspect its audio.
AUDIO / WAVEFORMWaiting for source
0:000:00
The multilingual Whisper model runs in your browser with WebGPU when available and falls back to WASM / CPU. The first run downloads and caches the model; your media remains on this device.

Manual alternative: play the source, move Start and End around one spoken line, type exactly what is said, then add the timed clip. Repeat for each line.

LOCAL BY DEFAULTWaveform analysis, speech recognition and WAV rendering happen on this device. The browser downloads the Whisper model, but it does not upload your selected media. Cloud transcription is not active.
PACK CLIPS

My first voice pack

00 selected / 00 detected

Your timed clips will appear here with duration and spoken content.

04 / PLAYABLE PACK CHECK

Finish inside the real game.

The installer writes validated files to the game's browser storage, refreshes Custom Voice Packs and selects the installed Pack. The Pack remains available on this device and browser.

Start the game when you have a Pack ready to test.
READY FOR A ROUND?

Take the Pack to the mic.

Open the main player in a new tab, allow microphone access, and start with the Pack you just installed.

Play now
LOCAL PARTY / UP TO FOUR

Bring the group together.

Gather the players around one screen, choose an order, pass the mic, and let every voice change the room.

Start a group game