I thought this research might be useful to people comparing speech components for Home Assistant Assist and other local voice setups.
While building Imprynt, I spent roughly a year collecting information about text-to-speech and speech-to-text systems. The lists became large enough that I decided to publish them as a free public resource:
- 168 TTS systems: Voices — text-to-speech compared · OpenGauntlet
- 106 STT systems: Listening — speech-to-text compared · OpenGauntlet
The entries cover local and hosted options, hardware requirements, licensing, languages and capabilities, lifecycle status, source information, and published benchmark data where available. This is a sourced research directory, not a claim that I personally benchmarked every system.
For Home Assistant users, I would especially appreciate feedback about missing local/offline engines, systems that work particularly well with Assist, outdated entries, or comparison fields that would make choosing a pipeline easier. Corrections and model requests can be submitted from the site, or posted in this thread.