BeforeAfterGirl voice changer, free
Five semitones up with the formants raised three. Lifting both by the same amount gives a chipmunk; lifting the formants less keeps it sounding like a person.
- +5 st
- formant +3
- high-pass 110 Hz
Free · record or upload · runs in your browser
Sample voice
0:30 · Girl / higher
Voice presets
Higher pitch, slightly smaller voice
Formant moves the voice's resonances (its “size”) separately from pitch. In a browser this is an approximation: big moves can sound hollow or phasey.
MP3 192 kbps or WAV 16-bit (320 kbps / 24-bit with Pro) · echo and reverb tails included · free, no watermark
Examples
Each clip is the same 12-second sung phrase from our sample voice, processed with this voice changer and nothing else. Press play, then flip Original / Changed: the switch keeps your place.
BeforeAfterFive semitones up with the formants raised three. Lifting both by the same amount gives a chipmunk; lifting the formants less keeps it sounding like a person.
BeforeAfterFive semitones down, formants down three and a little less treble. The harmonics crowd together, which is what makes a voice feel bigger.
BeforeAfterA 70 Hz ring modulator splits every harmonic into two, a 6-bit crusher adds grit and an 11 ms comb gives the metal ring.
BeforeAfterEverything outside roughly 400 Hz to 3.2 kHz is cut away, then a touch of drive and line hiss sells the phone.
BeforeAfterA 380 ms echo with long feedback inside a very large reverb. Pauses fill up with the tail of the last word.
BeforeAfterThe pitched harmonics are swapped for noise that follows the same mouth shape, so the words stay clear but nothing is sung.


Drag the handle. Pitch lifts the stack of harmonic lines; the formant move shifts where the bright bands sit. Same words, same timing, same length.

Press record, say a sentence, then click through the presets. Your take never leaves the browser.
How to
No account, no install. Works in current Chrome, Edge, Firefox and Safari on desktop and phone.
Press Record and allow the microphone, or drop an MP3, WAV, M4A or voice memo. The meter shows your level while you speak.
Choose one of 12 voices. Pitch, formant, echo and reverb sit right below; “More controls” opens the band, robot and space settings.
Flip Original / Changed at the same moment, then save an MP3 or WAV. The echo and reverb tails are rendered in.
Under the hood
A voice changer with effects is really two tools: one moves pitch and formants, the other colours the sound. These spectrograms are our own renders of the sample voice, before and after.

Pitch is how fast your vocal cords vibrate: the spacing of the harmonic lines. Formants are the resonances of your throat and mouth: the bright bands those lines pass through. Speeding a tape up moves both, which is the chipmunk sound. The girl and deeper presets move pitch by five semitones but formants by only three, so the voice changes register without turning into a cartoon.
Honest note: we move the formants by estimating the spectral envelope of every 23 ms slice and stretching it. That's an approximation, not a studio formant shifter; beyond about six semitones it starts to sound hollow.
BeforeAfterA ring modulator multiplies your voice by a steady tone. Every frequency f becomes two, f + 70 Hz and f − 70 Hz, so the harmonics stop lining up and the voice sounds machined. The robot adds a bit-crusher and a short comb; the alien uses a higher 420 Hz carrier, raised formants and a slow pitch wobble.
BeforeAfterClassic phone lines carried roughly 300 to 3,400 Hz, so those voices have no boom and no air. The presets cut both ends with steep filters, then add drive for the crunch of a small speaker and a little static. The megaphone keeps more top end, drives harder and adds a slap-back echo.
Pitch, formant and whisper are heavy, so they render in background workers a moment after you stop moving the slider and swap in where you are. Everything else is live Web Audio and responds as you drag. Export renders exactly the same chain.
Stage 1 · worker
Stage 2 · live
| Preset | Pitch | Formant | Band | Effects |
|---|---|---|---|---|
| Girl / higher | +5 st | +3 st | > 110 Hz | reverb 6% |
| Deeper | −5 st | −3 st | > 40 Hz | — |
| Chipmunk | +10 st | +10 st | > 160 Hz | — |
| Giant | −9 st | −7 st | > 30 Hz | reverb 20% |
| Robot | −1 st | −1 st | > 90 Hz | ring 70 Hz · 6-bit · drive 15% · echo 11 ms |
| Telephone | — | — | 400 Hz – 3.2 kHz | 10-bit · drive 25% · static 5% |
| Walkie-talkie | — | — | 500 Hz – 2.6 kHz | 8-bit · drive 55% · static 14% |
| Old radio | — | — | 220 Hz – 4.8 kHz | drive 30% · wobble 0.7 Hz · reverb 5% · static 8% |
| Megaphone | — | — | 600 Hz – 4.5 kHz | drive 70% · echo 110 ms · reverb 12% |
| Cave / hall | — | — | > 80 Hz | echo 380 ms · reverb 55% |
| Alien | +3 st | +7 st | — | ring 420 Hz · wobble 6 Hz · echo 60 ms · reverb 12% |
| Whisper | — | — | > 300 Hz | whisper 100% · reverb 8% |
Use cases
Short clips, narration and messages: anything you record first and share later.
robot · telephoneGive a sketch character its own voice, add a robot narrator or a phone-call cutaway, then drop the MP3 into your editor.
giant · walkie-talkieRecord lines for a mod, a tabletop session or a highlight edit: a giant, an alien, a radio call from the base.
whisper · deeperPosting a voice note in a public group? Whisper or deeper keeps the words and hides the obvious voice. It isn't forensic anonymity.
girl · cave · radioRead every part of a bedtime story or a radio play yourself, with a different voice for each character.
Use it kindly
A voice changer is for fun, characters and privacy. A few ground rules keep it that way.
Recording a friend? Get their OK before you change and share their voice, and respect a no.
Our presets are generic effects with no celebrity or character voices. Don't use them to pass yourself off as a real person, to trick someone or to get around a voice check.
In videos, podcasts and posts, a short “voice altered” note keeps your audience's trust.
Your recording is processed in the browser and never uploaded. A changed voice is harder to recognise by ear, but your words, accent and rhythm stay the same, and simple pitch changes can be partly reversed. For sensitive cases, combine whisper with a band preset and still assume it can be identified.
This voice changer works on clips you record or upload. It doesn't install a virtual microphone, so it can't change your voice in real time on Discord, Zoom or in games. Record, change, then share the file.
FAQ
Yes. Recording, all 12 presets, the custom controls and MP3 or WAV downloads are free, with no signup and no watermark.
Pick Girl / higher: pitch up five semitones, formants up three. Deep voices may need pitch +6 or +7 with formant +3 or +4. Raising formants as much as pitch gives a chipmunk.
No. It changes recordings and files, not a live microphone, and installs nothing. Record a clip, change it here, then share or upload the file.
No. Recording, processing and export all happen in your browser. The microphone switches off when you press stop, and closing the tab clears the take.
Moving a voice far from its natural range stretches the sound in ways a throat never would. Keep pitch within about ±7 semitones and the formant within ±4 for natural results.
MP3, WAV, M4A, AAC, OGG, Opus, FLAC and WebM up to 50 MB and 10 minutes long (500 MB and 2 hours with Pro), of which the first 5 minutes are used (10 with Pro). Microphone takes can be up to 1 minute, or 10 minutes with Pro. Output is a mono voice in a stereo MP3 or WAV.
Record a line, click through twelve voices, download the one you like. It takes about a minute.