self-hosted/ai
§01·model · /models

OpenAudio S1 Mini

ttsactiveCC-BY-NC-SA-4.0

Compact expressive text-to-speech model by Fish Audio, with voice cloning and emotion and tone control. The distilled build runs in roughly 5 GB, and that is what puts it on nineteen cards here at an 8 GB floor — the smallest tier this catalogue tracks — with both Radeon boards covered. Note the licence before shipping anything: CC-BY-NC-SA-4.0 is non-commercial.

§02·GPUs that run this model
19 total

benchmarked·~ runs via recipe (not benchmarked)· untested·doesn't fit