How old do you sound?
Your voice carries clues about you: how high it is, how fast you talk, how lively it sounds. Guess how well people read age from a voice, then let your device measure yours.
Read this sentence out loud
Tap the button and read at your normal speed. After 5 seconds the microphone switches off by itself, and your device measures three things about your voice.
On-device only: the sound is analysed in this tab with your browser's own audio tools. Nothing is saved or uploaded, and the microphone is switched off as soon as the 5 seconds end.
No microphone? Use sample voices
These are made-up voices: buzzing tones built by this page with a set pitch and speed. Not real people.
Why this matters: voices can be changed and copied
Voice-changing filters move your pitch, and the "size" of your voice, up or down. That's how an adult can sound like a child, or a young person like a grandparent.
Speech tends to slow down with age, and pitch shifts too. Researchers who asked people to disguise their age found they changed exactly these two things.
A voice-cloning tool learns someone's pitch, speed, accent and rhythm from a short recording. The copy sounds the same age as the real person, which is what makes fake family calls convincing.
If a call asks for money or codes, hang up and call back on a number you already have, or ask for your family's safe word.
Is that really your family calling?
Sources: Huckvale, M. and Webb, A., University College London, "A comparison of human and machine estimation of speaker age" (SLSP 2015): 36 listeners were off by 9.8 years on average (mean absolute error), and pooling 10 listeners brought it to about 7.4 years; the same paper lists earlier studies at 7.1 to 10.8 years (Braun and others 1999; Krauss and others 2002; Amilon and others 2007; Moyse and others 2014) and sums them up as "about 10 years". Skoog Waller, S. and Eriksson, M. (2016), "Vocal age disguise: the role of fundamental frequency and speech rate and its perceived effects", Frontiers in Psychology 7:1814: women's pitch changes little until menopause and then drops; men's pitch is lowest around 40 to 50 and rises again later; speech rate decreases with age for both. Wang and others (2023), "Neural codec language models are zero-shot text to speech synthesizers": a research model that copies a voice from a 3-second recording. How this page measures: pitch with the YIN method (de Cheveigné and Kawahara, 2002); speed by counting loudness peaks, which roughly follow syllables; liveliness as the spread of loudness in decibels. The age range on this page comes from simple made-up rules for illustration only. It is not a real age detector and not medical advice. Checked October 2026.