GoKawiilNetflix updates VMAF with new v1 model generation for video quality scoring
Netflix has released VMAF v1, a new generation of its open-source perceptual video quality models, alongside updated documentation comparing it to the prior v0 models. The release accompanies a technical blog post explaining the changes and guidance on selecting between model versions, and Netflix says it is seeking user feedback to refine the models further.
GoKawiil's interpretation of the reporting above, not reported fact.
VMAF is widely used across the streaming and codec industry to benchmark video compression quality, so changes to its scoring models could shift how engineers evaluate encoding performance. Netflix's openness to feedback suggests the v1 models may still be refined based on how the industry applies them in practice.
- Netflix released VMAF v1, a new set of perceptual video quality models.
- v0 models remain documented and supported alongside the new v1 release.
- Netflix is soliciting user feedback to further refine the VMAF v1 models.
NVIDIA Shield TV Pro — If you're geeking out over Netflix's VMAF perceptual quality metrics, you probably care about getting the best actual streaming video quality at home. The Shield TV Pro supports high-bitrate 4K HDR streaming and strong upscaling, letting you actually perceive the difference good encodes make. It's a great way to put VMAF-style quality improvements to a real-world test on your own TV.
See NVIDIA Shield TV Pro on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.Source: github.com, 2026-10-18
Published there as: “Netflix vmaf: Perceptual video quality assessment based on multi-method fusion”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.