A browser studio that turns a product photo and short brief into creator-style vertical clips using…
Download app now
Category
Sound
A broad audio section: lifelike speech generation, emotion analysis from the voice, music recognition, generated music and soundscapes, and conversational voice.
This is the catch-all audio category, so the tools differ more than the name suggests. Some generate speech, such as Aflorithmic, which offers natural voices with deep customisation, and Cartesia, built for very low latency. Some generate music. Others listen: Hume AI turns vocal and other signals into measurable emotional data, and SoundHound identifies music and searches lyrics.
Read more about this category
Latency matters only for live audio
If a voice will respond to a person in real time, as in a phone agent or a conversational assistant, delay is the first thing to measure, because a pause of a second already sounds wrong. For narration, podcasts or ads, latency is irrelevant; judge voice quality, pronunciation control and export formats instead. Aflorithmic and Speechify both sit on that side, one for producing voiced content and the other for reading documents aloud.
Music without a musician
Riffusion, Mubert and Soundful generate tracks from a prompt or a mood. Check the licensing terms before any of it goes into commercial work.
Once you know which side you are on, the narrower sections are more useful: Text to speech, Voice to text and Music.
