Text-to-Speech

Text-to-Speech is a product developed by MiniMax, the Chinese AI company founded in 2021 by Yan Junjie . In the broader market, open-source alternatives have emerged: Muyan-TTS from 沐言智语, released in mid-2025, claims the fastest inference among open models at 0.33 seconds per second of generated speech, with competitive word-error rates and voice quality scores . The technology is seeing applied use in consumer AI products—Me.bot's "Talks" feature, for instance, lets users distribute voice clones via H5 pages for asynchronous spoken communication, which founder Tao Fangbo argues lowers the barrier for emotional expression and human connection . Related coverage: MiniMax's own voice model updates , Google I/O 2025 .