Hacker News Viewer

Inflect-Micro-v2: complete voice in 9.36M parameters

by nateb2022 on 7/26/2026, 12:36:30 AM

https://huggingface.co/owensong/Inflect-Micro-v2

Comments

by: yjftsjthsd-h

Couple highlights:<p>&gt; Complete local text-to-waveform speech synthesis under 10M parameters.<p>In case, like me, you hoped &quot;complete&quot; voice might mean <i>both</i> stt and tts. Not to speak poorly of it, just clarifying.<p>&gt; English only, with one fixed male voice. This is not zero-shot voice cloning.<p>(And then a bunch of statements on limitations that I read as &#x27;quality can be spotty but if you play with it it should be fine&#x27;) But like. In &lt;10M params I&#x27;m not judging:)

7/26/2026, 3:21:08 AM


by: tmaly

This is impressive. I wish there were a voice clone option.

7/26/2026, 2:01:19 AM


by: jsomedon

amazing quality for such small size!

7/26/2026, 2:55:44 AM