US firm Resemble AI open-sourced its voice-cloning platform 'Chatterbox', capable of text-to-speech conversion and rapid voice cloning. Chatterbox can mimic emotions ranging from monotone to highly expressive. In a comparison with ElevenLabs, 63.75% of listeners preferred Chatterbox. The technology underpins the growing market for voice-cloning, including MIT and Tsinghua University's OpenVoice and Zyphra's model that needs just five seconds of sample audio.