Meta Muse Video: What We Know So Far

Meta Muse Video is a video generation model from Meta Superintelligence Labs. Meta previewed it on July 7, 2026, together with the launch of Muse Image. Here are the facts Meta has confirmed:

• Who made it: Meta Superintelligence Labs, as one of its first media generation models.
• What it does: Generates video with sound. Native audio is part of the model.
• Built on: The same pretraining base as Muse Image.
• Strengths Meta names: Prompt adherence, visual fidelity, and temporal consistency.
• Arena rank: No. 3 for text-to-video in human-preference Elo, as of July 5, 2026.
• Known gaps: Audio and video synchronization, and physically accurate fast motion.
• Status: An early preview, "coming soon" to creators and Meta AI.

Want to try it now? Generate a video with Muse Video.

What Meta says. In its July 7, 2026 post, Meta describes Muse Video as an early preview and says it is coming soon to creators and in Meta AI. Meta has not named which apps or countries will get it first.

Why you may see different answers. Some articles say Muse Video is already in the Meta AI app, on meta.ai, in Instagram Stories, and in WhatsApp. Those are the places where Muse Image launched. Meta's post lists those places for Muse Image, not for Muse Video. If you read a claim that Muse Video is live in those apps, check it against Meta's post.

Where you can use it today. You can use the Muse Video model on Muse-Video.app. Write a prompt, or upload an image, and generate a short clip with sound.

How to tell which is which. Meta's own products live on Meta domains and apps. Muse-Video.app is a separate product powered by the model. If a site says it is "official Meta Muse Video", look for a link from Meta before you pay.

Meta says Muse Video is "coming soon". It has not given a date, so we do not either. We will update this table when Meta does.

Meta says Muse Video is competitive in three areas.

Prompt adherence

Does the clip show what you wrote? That means the subject, the action, the place, and the camera move. Meta lists this as a strength.

Visual fidelity

Detail, light, texture, and depth. Meta's preview clips are meant to show this. As with any model, judge it on whole clips, not on one good frame.

Temporal consistency

People, objects, and backgrounds should stay the same from the first frame to the last. Meta lists this as a strength too.

Native audio

Sound is generated with the video. You can describe the sounds you want in the prompt, such as footsteps, a kettle, or a short line of speech. See our prompt guide.

Meta is open about two gaps.

• Audio and video sync. Sound and picture do not always line up exactly. Do not expect perfect lip sync or exact sound timing.
• Fast physical motion. Quick collisions, sports, and other fast, physical movement are harder to get right.

Plan around these. Keep early tests short, check speech and sharp sounds before you use a clip, and generate more than one version.

Meta reports that Muse Video ranks No. 3 for text-to-video on Arena, based on human-preference Elo, as of July 5, 2026.

What that means: Arena asks people to compare two clips and pick the one they prefer. A rank on that board shows how people rated the model, not how it does on every task. Rankings also move when new models arrive, so check the current board before you quote the number.

How it compares with other models: see Muse Video vs Seedance 2.5 and Muse Video vs Kling 4.0.

If you want the image model, it is the one that is live. This page is about the video model.

Meta uses Content Seal, an invisible watermark, on images made with Muse Image in the Meta AI app and on meta.ai. Meta says the mark survives cropping, compression, resizing, and screenshots. For video, Meta says Content Seal is not applied yet and that it plans to extend it to video soon.

If you share AI video, say that it is AI-generated. Check the rules of the platform where you post.

Option 1 — Wait for Meta's rollout. Meta says Muse Video is coming to creators and Meta AI. However, you'll need to join the waiting list.

Option 2 — Use it on Muse-Video.app today.

1. Write one or two sentences: the subject, the action, and the place.
2. Add a camera move and the sounds you want.
3. Generate. Change one detail and try again if the result is off.

Start here: Muse Video AI Video Generator · Muse Video Prompt Guide