AI VOICE DECOY
Your voice.
Decoy speech.
Read a short passage aloud. Create fictional speech in your voice.
Connect a speaker, press Play, and choose new stories or repeat.
Record your voice
Mic offRead this in your usual voice.
Press Record, then read the passage below aloud. Speak naturally in a quiet room for about 20 seconds.
Speak at a comfortable pace; you do not need to finish every word. We check the audio level and keep up to 12 seconds for the voice model.
Read this aloud in your usual voice:
Recording stops after 20 seconds. Keep speaking naturally, with small pauses between sentences.
Check your recording
This is the part of your recording the voice model will use. Listen for clear speech, without music or another person talking.
My voice
Saved in this browser
Hear a short new sentence in your generated voice. Compare its tone and accent with your original recording.
The first preview downloads about 180 MB. Keep this page open while it prepares.
Playback
Synthetic voiceYOUR VOICE · FICTIONAL SPEECH
Ready when you are.
Record your voice to get started.
Preparing the voice engine…
Uses the speaker connected in your device’s Bluetooth or audio settings.
Leave this page visible on the phone.
Choose what to say
CONTENT & TIMINGCreate new passages automatically. The current passage repeats while the next one is preparing.
Up to 4,000 characters. Choose Play one passage or Repeat this story. Your text stays in this tab unless you save its generated audio to your library.
Original story templates are shuffled on your device so each plot is used before it repeats within a category. Short passages are complete stories; longer passages add detail. The voice model creates the spoken audio.
Prepare a passage without playing it, then save it for another visit. You can also save the passage that is currently playing.
Saved audio
PREPARE ONCE · PLAY AGAINReady-to-play passages 0 / 20 passages · 0.0 / 64 MB
Saved passages play without starting the voice model. Repeat uses the playback time you choose above. Your saved audio stays in this browser; clearing site data or browser storage cleanup can remove it. Download a WAV for a separate copy.
Your library is empty. Prepare or play a passage, then choose Save this audio.
First-time setup
Recording needs microphone permission. Read at your usual pace for about 20 seconds; the app checks the recording and keeps up to 12 seconds. The level and clipping checks cannot tell whether the sound is speech, music or background noise, so listen to the selected recording yourself.
Choose Try my generated voice for a short audition of new words. The first preview or Play downloads about 180 MB of voice-model files. Use Wi-Fi and keep the tab open. Models are cached when your browser allows it. Short passages need less speech to be prepared. Replaying the last generated audio does not need the voice model to start again.
Voice likeness and speed vary by device and recording. Each browser has its own voice recording; it is not synced to other devices.
Repeat, change & leave playing
Pair a Bluetooth speaker or connect a wired speaker in your phone or computer settings. Return here and choose Check speaker, then Play. Start with a comfortable volume.
Repeat this story makes the audio once and loops it. Keep making new stories prepares fresh passages in your voice and changes at the end of a passage. If preparation takes longer, the current passage repeats until the next is ready. If the next generation fails, the current passage keeps playing and a message explains what happened. Press Next to try again.
Mix it up draws from stories, everyday thoughts and mini talks. Choose a playback time or Until I press Stop. The timer counts audio playback; pausing and initial preparation do not use up the time. Volume, content category and playback choices are remembered when browser storage is available. Playback never starts automatically on a new visit.
Prepare audio silently creates one passage without playing it. Choose Save this audio, then use its Play button in Saved audio whenever you return. Up to 20 passages or 64 MB of audio can be saved; the browser may have less space available. Saved audio can play even when this device cannot start the voice model.
Leave the phone with the speaker and keep this page visible. Screen awake and supported headset or system media controls help you manage playback, but the phone may refuse or release them. Locked-screen and background playback or generation are not guaranteed. Download a WAV to use your usual audio player.
Your recording & data
Your voice sample and its derived voice data are kept in this browser if storage is available. Saved passages include their audio, title, text and creation date. They are not uploaded by this app. On a shared device, another person using the same browser profile may be able to access them.
Replacing your voice keeps previously saved passages. Delete voice & saved audio removes the app’s voice and all saved passages, and clears the running voice engine. Downloaded WAV files are separate and must be deleted from your Downloads folder yourself.
The app downloads its code and voice models from this site. It does not monitor the room during playback. Generated audio is synthetic. This app does not detect listening devices or provide verified protection against recording or audio analysis.