aaronfeng
patreon
[News] Double the Speed, Keep the Quality: Omni-Turbo Model
🕑 Added 2025-08-25 14:39:04 +0000 UTCIn the NEXT update, a Turbo version of Omni-Small will be available.
This Turbo version aims to preserve the quality of the Omni-Small model while boosting speed.
Speed:
The turbo model nearly doubled the speed of Omni-Small, while still faster than the quality mode of RTX-SuperRes
Quality:
The de-noise capability of Turbo is significantly better than Mini. But when compared to Omni-Small, Turbo's result is less sharp when zoomed in.
🖼️Compare results on Imgsli:
This is only the first version of the Omni-Turbo model. I'm still working on future updates for this model. Thank you for supporting this project! 💕
