Most video comparisons in Vizify start with audio, because it varies so much between models. Here it does not: Veo 3.1 and MiniMax H3 both render native audio on every clip, and neither offers a silent option. If a silent result is a requirement, look at Kling 3 vs Seedance 2.5 instead.
The picture itself is where they separate. Veo 3.1 returns an 8-second clip at 720p, framed Auto, 16:9, or 9:16. MiniMax H3 returns 2K, lets you choose any length from 4 to 15 seconds, and adds square and 21:9 framing — though not Auto.
References work differently too. Both accept up to two images, but MiniMax H3 treats them as first and last frames by default, and it also accepts short video and audio reference clips; adding one switches the images to identity or style guidance. Veo 3.1 takes images only. MiniMax H3 is also the successor to Hailuo 2.3 in Vizify, which stays available on its own page.