See our contact page for ways to get in touch.

RSS: Subscribe to the RSS feeds here

MIMIC3 is an open source and privacy focused text speech engine. It can be called over a network if you'd like to have a beefy central system to run it, or it'll run on hardware as light as a pi four without strain. I found the voices for MIMIC3 are good and don't sound as robotic as most other TTS engines. There should be some way of reproducing a model and then reproducing from the same seed, the output. Who knows where we'll be by next week?