Your voice.
Every document.
Record yourself for a minute, then let anything you have written speak in your own voice — notes, drafts, papers, whole books.
Free to start. No credit card.
The rain had stopped by morning. She opened the notebook to the first page and began to read it aloud. Every sentence sounded familiar — it was her own voice, reading her own words back to her.
Press play to hear this page read aloud. This sample uses the engine's built-in voice — your documents are read in your own cloned voice.
From a page to a voice in three moves.
-
Teach it your voice
A clean recording of one minute is enough. You type the exact words you spoke so the model can line the sound up with the text. Keep as many voices as you like — your reading voice, your presenting voice, someone else's with their permission.
-
Hand it your text
Upload a plain .txt file and choose a voice. Long documents are fine; the job runs in the background and you can close the tab.
-
Listen and decide
Finished audio appears in your workspace with a player. Keep it, download it, or delete it — the file goes with it.
Listen to the engine
Short passages in Chinese and English, read start to finish with natural pacing.
-
A bedtime story
小镇的尽头有一座旧灯塔。每天傍晚，守塔的老人都会提着一盏小灯，慢慢走上长长的台阶。有一天，一个迷路的小男孩坐在灯塔下面哭。老人没有多问，只是递给他一杯热牛奶，轻声说：“别怕，灯亮了，船就能回家，你也能。”那天晚上，灯塔的光照得很远很远。
-
A short article
Most of us read with our eyes, but people listened to stories long before they read them. A voice adds pace, warmth and small pauses that help a sentence land. That is why a chapter read aloud can feel easier to follow on a long walk or a quiet evening, when your eyes need a rest but your mind does not.
-
A letter home
亲爱的妈妈：这几天降温了，记得多穿一件外套。我在这边一切都好，工作虽然忙，但每天都能按时吃饭。周末我去了一趟海边，风有点大，海水却很蓝。等忙完这一阵，我就回家看你，到时候我们一起去公园散散步。
-
A slower reading
Every evening, my grandmother waters the plants on her balcony. She talks to them softly, as if they were old friends. When I asked her why, she smiled and said, "Everything grows better when someone is listening."
These samples are read by the engine's built-in voice, not a cloned person. Your documents are read in your voice, cloned from your own recording.
Built for people who write a lot.
- Several voices per account
Switch between them per document.
- Chinese and English
The voice model handles both, and mixes them inside one document.
- Nothing to install
Recording happens in the browser. Playback too.
- Queue you can watch
Every job shows its status while it runs, so you know what is waiting on what.
- Yours to delete
Remove a voice or an audio file at any time and it is gone from storage.
Where your voice lives.
Your reference recordings and finished audio are stored in the operator's Google Drive storage. MyVoiceTale keeps only your email, your password hash, and the list of what you have made. Delete a voice or a file and the stored file is removed too.
- In the operator's storage
- Reference recordings and generated audio.
- On our side
- Only your email, your password hash, and the list of what you have made.
- When you delete
- The stored file is removed too.
Questions people ask first.
How much audio do you need?
Between ten and sixty seconds of clear speech, saved as a WAV file. A phone recording in a quiet room works.
What can it read?
Plain text files up to 200 KB, in Chinese, English, or a mix.
Can I clone somebody else's voice?
Only with their knowledge and consent. Accounts used to impersonate people are removed.
What happens to a document after it is read?
The text file and the audio are both kept in the operator's storage until you delete them.