Audio

Speech-to-speech & voice changing

Convert one voice into another while keeping timing and emotion.

What it is

Speech-to-speech (voice changer) APIs take a recording of someone speaking and re-render it in a different voice. Unlike text-to-speech, the original performance (pacing, emphasis, laughter, emotion) is kept, so you get acting control that typed text cannot give.

How pricing works

Billed per minute of input audio (or by credits that map to minutes). We normalise to one minute of audio converted. Custom or cloned target voices may add a monthly or per-voice fee that is not in this number.

We compare everything in one unit: USD per minute of audio.

Typical uses

Voice acting and character voices for games and animation
Anonymising speakers while keeping natural delivery
Dubbing a performance into a stock or cloned voice
Live voice conversion for streaming and calls

What to compare

Naturalness and how well emotion is preserved
Similarity to the target voice
Latency for real-time use
Robustness to noisy source audio

Watch out for

Quality depends heavily on the source recording; noisy input gives artefacts.
Cloning a real person's voice needs their consent, and many providers require verification.
Real-time conversion is often lower quality than offline processing.

Capabilities explained

Real-time conversion: Converts live audio with low delay, for streaming or calls, instead of only processing finished files.
Custom target voice: You can convert into a voice you cloned or trained yourself, not only stock voices.
Keeps emotion and timing: The conversion keeps the original pauses, emphasis and emotion rather than flattening them.
Languages: Number of languages the service handles. More is better only if your target languages are on the list; check them individually.
Typical latency · ms: How long until the first useful output arrives. Lower is better for interactive apps; batch jobs rarely care.
Commercial use allowed: Whether outputs may be used in a commercial product under the provider's terms. Always re-check the license for your case.

Glossary

Voice conversion: Changing who a recording sounds like without changing what is said or how it is performed.
Target voice: The voice the audio is converted into.

All 2 APIs, cheapest first

Free to browse

Tick 2 to 4 APIs to compare them side by side.

API	$ / audio min	Real-time conversion	Custom target voice	Keeps emotion and timing	Quality	Price data	Compare
ElevenLabs Voice Changer ElevenLabs	$0.12	No	Yes	Yes	90	Verified Oct 6, 2026 · aging
Kits.AI voice conversion Kits.AI	$0.15	No	Yes	Yes	–	Unverified sample