The newly launched Nero Start application incorporates extensive user feedback to deliver a streamlined and adaptable application launch experience.
With Nero Start new mode, now you can:
Discover the new mobile application: 
Erasee – Remove Background in 5 Seconds. Try it now: https://pcai.nero.com/erasee .
Support AI subtitle functionality to recognize speech and convert to subtitle automatically by online service. It also provides manual subtitle editing functionality including edit timecode and text style.
PS. This option might overwrite your original files when you export the results to the same directory as the original files.
The new “Face enhancement” model has been added to Nero AI Video Upscaler since version 27.0.1.7 & 27.5.2.12
The “Face Enhancement” model is suitable for improving the details of a person’s facial contours. This model is suitable for real video footage and is not suitable for faces in CG or animation.
See detailed comparison with other models. Compared to the Realistic and Versatile models, the Face Enhancement model balances facial clarity and smoothness.
Note: the “Versatile” model only available for the Nero AI Video Upscaler subscription version. Windows Store Pro version will get the model updates later.
Please check the following preview screenshots and recommendations are on the below.
All the original videos resolution are 1280×720. Upscale to 3840×2160 (4K) and zoom to 200% to preview.
Animation model is strongly recommended to increase the sharpness and contrast to get the clear view of the upscaled video.
Versatile model is recommended to keep the details of the animals. If you feel the processing time is too long, Fast model can also be used but it keeps more noises than the versatile model.
Realistic model is recommended to refine the skin texture and suppress the noises. If you prefer to keep the skin texture, versatile model is also recommended.
Realistic model is recommended to refine the edges by sharpening the video and suppress the noises. If you prefer the less sharpness view, versatile model is also recommended.
Versatile model is strongly recommended to keep the sharpness and slightly increase the contrast to reserve more details from the original video.
Realistic and versatile models are both recommended to retain the sharpness and reduce the blurry. If the original video has been applied some post-processing effects like film grain, you can use the versatile model to keep these effects. If it is not, realistic model can process the video more faster.