Magnific's Speak now defaults to VEED Lipsync 2.0
By Steven Van ·
The new default model dubs up to 10 minutes of real or AI-generated footage at 4K, preserving performance while swapping language.
Magnific added a best-in-class lip-sync engine to its video toolkit: Veed Lipsync 2.0 by VEED is now on the platform, and it's already the default model in Speak, Magnific's talking-video feature. The pitch is dubbing that doesn't flatten the delivery, localize while keeping the performance intact.
Real footage and AI video, at length
Two specs make this production-grade rather than demo-grade. It works on both real footage and AI-generated video, so you can relip an actual filmed interview, not just synthetic clips. And it handles up to 10 minutes in 4K, well past the few-seconds ceiling most lip-sync tools hit. That combination, long-form, high-res, real footage, is what a dubbing job for an actual video actually requires.
Keeping the performance is the hard part
Naive lip-sync solves the mouth and breaks everything else, the emotion drains out, the timing goes robotic, and the result reads as dubbed. VEED's Lipsync 2.0 is built to preserve the performance while swapping the language, which is the difference between a localization your audience accepts and one they click away from. Making it the default in Speak signals Magnific's confidence that it's the version to reach for.
VEED's model is everywhere this week
The same Lipsync 2.0 landed on fal the same day as a developer API, which tells you where lip-sync is heading: a specialized model distributed across every platform, with each host wrapping it for its audience. Magnific's angle is in-app and creator-facing (it's the engine behind Speak); fal's is programmatic. Either way, high-quality dubbing is becoming a commodity capability, and this week it got noticeably better and more available at once, alongside OpusClip and HeyGen shipping their own dubbing tools.
Try it
Veed Lipsync 2.0 is live on Magnific now and powers Speak by default. The real test is a filmed clip with genuine emotion, dub it and watch whether the feeling survives the language swap.
Sources: Magnific on X, VEED.