Tarmac AI voice generator & voice changer
Raspy, blunt and secretly soft. Type text and hear it in Tarmac's voice (text to speech), record or upload your own voice to change it, or make a song cover. Free, no signup.
- 1Pick a voice
- 2Record, upload a file or song, or type text
- 3Listen, download and share
Auto pitch moves your voice into Tarmac's range. If it sounds too high or low, nudge the slider (12 = one octave).
Pick a voice
About the Tarmac voice
Tarmac is an original AI voice from the voicechange.io catalog, not modeled on any real person, show, game or film. Built as a character of its own, Tarmac is the kind of voice you picture belonging to someone who has spent a lot of years on the road and has stories to tell. It is blunt, a little worn around the edges and secretly soft underneath.
Expect a raspy, grainy tone with a lower, grounded pitch. Tarmac speaks at an unhurried, steady pace and says things plainly, without dressing them up. Sentences land with weight, and the roughness gives even simple lines a lived-in feel. Every so often a gentler warmth shows through, which is where the secretly soft side comes in. It suits gruff narration, dry one-liners, character dialogue and songs with a bit of grit.
To get good results on voicechange.io, you can record or upload speech and hear it turned into Tarmac, type text and listen to Tarmac speak it, or make a song cover. The pitch is matched to the voice automatically, so there is nothing to adjust. For speech, use a clear recording with little background noise and speak in a relaxed, steady way. Short sentences and natural pauses play to this voice, and a calm, low delivery brings out the rasp best.
Voice details
| Category | Original voices |
|---|---|
| Model type | RVC voice conversion model (fan-made) |
| Centre pitch | About 166 Hz, a mid-range voice |
| Works with | Recordings, uploads, text to speech and song covers |
| Last updated | November 16, 2026 |
Tips for the best Tarmac result
This model is centred around 166 Hz. With auto pitch on, a typical deeper voice (around 120 Hz) is shifted about 6 semitones up, and a typical higher voice (around 210 Hz) is shifted about 4 semitones down. Gaps under 4 semitones are left alone, and shifts stop at one octave. If the result sounds strained, nudge the slider one or two steps back toward your own voice; for song covers, keep it at 0, +12 or −12 so the vocal stays in key. How pitch works.
- Record in a quiet room with no music in the background; noise is the main cause of robotic-sounding results.
- Perform the line the way Tarmac would say it. The model copies your timing and emotion, only the voice changes.
- For song covers, pick a track with one clear lead vocal and little reverb.
This is a fan-made AI voice model. voicechange.io is not affiliated with or endorsed by Tarmac, the voice actors or the rights holders. Please label shared results as AI and don't use them to mislead anyone. Rights holders can request removal on the contact page.
Tarmac voice questions
Is the Tarmac AI voice free?
Yes. You can use the Tarmac voice on voicechange.io for free with no account: type text, change a recording or make a song cover. Free downloads keep up to 3 minutes per clip.
Can I use the Tarmac voice live on Discord?
Not this model: website voices convert recordings and text, which takes 10 to 60 seconds. For a live voice in Discord, games or Zoom, use the free VoiceChange desktop app, which runs its own voices in real time.
How do I make my voice sound more like Tarmac?
The model is centred around 166 Hz, a mid-range voice. Leave auto pitch on so your recording is shifted into that range, record in a quiet room, and copy how Tarmac speaks: the model keeps your timing and emotion and changes only the voice.
Is this the official Tarmac voice?
No. It is a fan-made AI voice model and is not affiliated with or endorsed by Tarmac, the voice actors or the rights holders. Please label anything you share as AI.
Want to talk in a new voice live? Get the free desktop app for Discord and games.