Tech Current

Tag: speech-synthesis

Inflect-Micro-v2 packs local speech synthesis into 9.36M parameters
SoftwareJul 26, 2026

Inflect-Micro-v2 packs local speech synthesis into 9.36M parameters

A new open-weight text-to-waveform model from Owen Song aims to deliver complete English voice synthesis under 10 million parameters, with CPU and CUDA support and deterministic seeds.

text-to-speechspeech synthesisopen-weightONNX

Analytics, advertising, and privacy choices

We use analytics and advertising cookies only with your permission. You can change this choice later from the footer.

Privacy policy