Seedance 2.0 Mini is officially live with lower pricing.

Create with Muse Video from Meta Superintelligence Labs

Muse Video is Meta's next-generation text-to-video model — built on the same pretraining base as Muse Image, with exceptional visual fidelity, native audio support, and strong prompt adherence. Preview what's coming to creators and Meta AI.

0/2500
Duration
4s
Estimated Cost48 credits

Muse Video combines Meta Superintelligence Labs' agentic media stack with competitive text-to-video performance — ranking among top models on human-preference benchmarks.

Why creators choose Muse Video

Exceptional visual fidelity

Muse Video delivers sharp, cinematic visuals with strong detail retention across frames — built on the same foundation as Muse Image for consistent quality.

Native audio generation

Generate video with synchronized native audio in a single workflow. Describe speech, ambience, or music in your prompt and let Muse Video align sound with motion.

Strong prompt adherence

Strong prompt adherence

Muse Video follows complex creative instructions faithfully — from camera direction and scene composition to character actions and pacing.

Temporal consistency

Temporal consistency

Maintain coherent motion, identity, and environment across shots. Muse Video is designed for multi-beat storytelling with stable visual continuity.

Built on agentic Meta AI research

Built on agentic Meta AI research

Like Muse Image, Muse Video inherits Meta's agentic approach — reasoning, tool use, and self-refinement at inference time for higher-quality generations.

Use the video generator below to experiment with AI video creation while Muse Video rolls out to creators and Meta AI.

How to create with Muse Video

Scroll to the generator section and describe your scene — subject, camera movement, lighting, dialogue, and audio cues.

Open the video generator

Muse Video is designed for creators, filmmakers, and teams who want Meta-grade visual quality with native audio in their AI video workflows.

Who is Muse Video for?

Social creators and marketers

Social creators and marketers

Produce short-form video ads, product teasers, and social content with cinematic visuals and synchronized audio — ready for Instagram, Reels, and campaigns.

Filmmakers and directors

Filmmakers and directors

Pre-visualize scenes, test camera language, and explore multi-shot sequences with strong temporal consistency before full production.

Brands and creative teams

Brands and creative teams

Prototype branded video concepts quickly, align stakeholders on look and feel, and scale content across Meta's ecosystem when Muse Video launches broadly.

Start creating AI video today

Experiment with our video generator now — and stay tuned as Muse Video comes soon to creators and Meta AI with even stronger fidelity and native audio.

Key strengths from Meta Superintelligence Labs' early Muse Video preview and research blog.

Muse Video capabilities

Text-to-video generation

Turn detailed prompts into dynamic video clips with competitive performance on prompt adherence and visual quality benchmarks.

Native audio in one pass

Muse Video supports native audio generation alongside visuals — reducing the need for separate sound design passes on supported workflows.

Shared foundation with Muse Image

Built on the same pretraining base as Muse Image, enabling a unified agentic media stack across image and video creation in Meta AI.

Arena-ranked performance

At launch preview, Muse Video ranked No. 3 in human-preference Elo for text-to-video on Arena — competitive with leading commercial models.

Creator and Meta AI rollout

Muse Video is coming soon to creators and Meta AI — following Muse Image's launch across Meta AI app, meta.ai, Instagram Stories, and WhatsApp.

Content Seal provenance (video soon)

Meta plans to extend Content Seal invisible watermarking to Muse Video — helping verify AI-generated media provenance across the Meta ecosystem.

Frequently asked questions

Still have questions? We're here to help.