Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
ReleaseAPI / Product · ActionableThe Decoder→ original

Suno adds lip-synced music clips of up to 15 seconds to its AI songs

Suno has launched Visuals: a song made by its AI can now get a video clip of up to 15 seconds. Previously, creators had to go to a separate video service and explain all over again what the track was about. Now the clip follows the song's structure, picks up the words and syncs the character's lips with the vocals. You can start from a photo, a text prompt or a ready-made Suno character. For anyone making Shorts and Reels, it is a path from idea to clip inside a single app.

AI-processed from The Decoder; edited by Hamidun News
Suno adds lip-synced music clips of up to 15 seconds to its AI songs
Source: The Decoder. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceThe Decoder ↗
Editor-in-Chief
Key metric #1H3 Move — MiniMax's H3 video model, customized for Suno
Key metric #2Two quality tiers, Fast and Standard; separate credits are required
Key metric #3Filters block prompts that mention artists, celebrities and protected works

The Suno music generator is no longer limited to audio. The service has launched Visuals, a feature that creates images and video clips of up to 15 seconds for a song made by its AI. The problem will be familiar to anyone who posts tracks on social media: the song is finished, but there is nothing to package it with.

Until now, you had to open a separate video service, describe the mood again and fit the visuals to the rhythm by hand. Now, for example, a creator uploads their own photo, and the character on screen moves their mouth in time with the vocals and changes the scene on the chorus. This matters to people who make content for Shorts, Reels and streaming platforms: the whole path from idea to clip happens in one app.

How clips work in Visuals

The feature runs on H3 Move. This is a customized version of H3, a video model from the Chinese company MiniMax, with safety filters added. Judging by the description, Suno did not train a video model from scratch. It took an existing one and adapted it to its own task.

Clips run up to 15 seconds. They follow the structure of the song, so the verse and the chorus look different. The model picks up the words from the lyrics and supports lip sync: the character's lip movements match the vocals. That is the difference between "a pretty picture set to music" and a video in which someone is actually singing.

There are three ways to start: upload a reference photo, write a text prompt or choose one of Suno's ready-made characters. The last option suits people who have no face or persona of their own for a channel.

There is also a separate lyric video mode, a clip that displays the song's lyrics. In it, you can change fonts and colors, extend clips on a timeline, and choose the duration and aspect ratio. The format matters in practice: vertical video for Shorts and horizontal video for YouTube need different layouts.

Pricing, limits and safeguards

Two quality tiers are available: "Fast" and "Standard." It is logical to assume that the first suits quick drafts and the second a more polished result, but the source does not describe the difference in detail. Visuals works on the web, on iOS and on Android. The feature uses its own credits, separate from the ones spent on music. Pro and Premier subscribers get a temporary allowance to try it.

Suno built in safeguards up front. Filters block prompts and uploads that mention specific artists, celebrities or copyrighted works. A watermark is embedded in the clip before it is published outside Suno. The source does not say whether the watermark is visible to the eye or exactly how it will be checked.

Why the company is being so careful

The context explains the caution. In late July, the Munich I Regional Court ruled against Suno in a suit brought by GEMA, the German music rights organization, and found copyright infringement. Soon afterward, Suno announced tools that mark songs made on the service as AI-generated on other platforms. The company has an agreement with Warner Music, while Universal and Sony are continuing their lawsuits.

Against this backdrop, the ban on mentioning artists and the watermark look like a natural step. A video feature with recognizable faces and other people's likenesses would give opponents one more reason to file new claims.

What this means for practitioners

For product managers and founders, there are two takeaways. First, AI music services are turning into full content-production kits. The user gets a song and a clip in one place, and has less and less reason to switch to other tools. Second, Suno took someone else's video model and adapted it for itself. For small teams, that is a workable strategy: you do not have to train your own model to ship a video feature.

There are limits, too. A 15-second clip does not replace a full three-minute video sequence; you would have to assemble one from pieces on the timeline. The filters restrict creators of covers and fan content. Separate credits mean that a song with a video will cost more than a single track, but the announcement does not say how much more.

The direction is clear, though: the generative music market is moving toward a "sound plus video" bundle, and the legal risks in it will be managed with the same filters and labels as in the music itself.

Story

Key Questions Answered

How clips work in Visuals?

The feature runs on H3 Move. This is a customized version of H3, a video model from the Chinese company MiniMax, with safety filters added. Judging by the description, Suno did not train a video model from scratch. It took an existing one and adapted it to its own task. Clips run up to 15 seconds. They follow the structure of the song, so the verse and the chorus look different. The model picks up the words from the lyrics and supports lip sync: the character's lip movements match the vocals.

What are the key details on Pricing, limits and safeguards?

Two quality tiers are available: "Fast" and "Standard." It is logical to assume that the first suits quick drafts and the second a more polished result, but the source does not describe the difference in detail. Visuals works on the web, on iOS and on Android. The feature uses its own credits, separate from the ones spent on music. Pro and Premier subscribers get a temporary allowance to try it. Suno built in safeguards up front.

Why the company is being so careful?

The context explains the caution. In late July, the Munich I Regional Court ruled against Suno in a suit brought by GEMA, the German music rights organization, and found copyright infringement. Soon afterward, Suno announced tools that mark songs made on the service as AI-generated on other platforms. The company has an agreement with Warner Music, while Universal and Sony are continuing their lawsuits. Against this backdrop, the ban on mentioning artists and the watermark look like a natural step.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Private AI Builders Community · Vibe Club

Ship AI Faster with Founders, Engineers & Weekly Architecture Reviews

Inside Vibe Club (hamidun.com/club) — weekly live masterminds with Zhemal Khamidun, Tech Radar, full LMS access, and 370+ production Claude Code skills.

Join Vibe Club →Explore Hamidun Academy
What do you think?
Loading comments…