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.
Muse Video AI Generator
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.
Muse Video
74 / 2500
Resolution
Aspect Ratio
Video Length
Mode
Estimated Cost 160 credits
40 credits/s · 480p
Output
Result
Your Muse Video will appear here
Submit a prompt to create a Muse Video generation task.
Features
Why creators choose Muse Video
Muse Video combines Meta Superintelligence Labs' agentic media stack with competitive text-to-video performance — ranking among top models on human-preference benchmarks.
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
Muse Video follows complex creative instructions faithfully — from camera direction and scene composition to character actions and pacing.
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
Like Muse Image, Muse Video inherits Meta's agentic approach — reasoning, tool use, and self-refinement at inference time for higher-quality generations.
Workflow
How to create with Muse Video
Use the video generator below to experiment with AI video creation while Muse Video rolls out to creators and Meta AI.
1
Open the video generator
Scroll to the generator section and describe your scene — subject, camera movement, lighting, dialogue, and audio cues.
2
Add references if needed
Upload a first frame or reference images to guide identity, style, or composition — then refine your prompt for motion and pacing.
3
Generate and iterate
Generate your clip, review the result, and iterate on prompts or references until the output matches your creative vision.
Audience
Who is Muse Video for?
Muse Video is designed for creators, filmmakers, and teams who want Meta-grade visual quality with native audio in their AI video workflows.
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
Pre-visualize scenes, test camera language, and explore multi-shot sequences with strong temporal consistency before full production.
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.
Try Muse Video
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.
Generate a video
Capabilities
Muse Video capabilities
Key strengths from Meta Superintelligence Labs' early Muse Video preview and research blog.
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.
Models
Explore media family
Muse Video sits alongside Muse Image, Muse Spark, and Meta AI — one connected stack for agentic image and video generation.
Muse Video
Meta's preview text-to-video model with exceptional visual fidelity, native audio, and strong temporal consistency for creator workflows.
Muse Image
Agentic image generation with tool use, search, self-refinement, and precision editing — available today across Meta AI products.
Meta AI
Create, edit, and reimagine media with friends across Meta AI app, meta.ai, Instagram, WhatsApp, and Facebook.
Muse Spark
Integrated with Muse Image for joint planning, code execution, and animated media — powering agentic creative workflows.
Learn
Learn more about Muse Video and Meta AI
Read Meta Superintelligence Labs' research announcement, benchmark results, and product roadmap for Muse media generation.
Introducing Muse Image and Muse Video
Muse Video Arena rankings and text-to-video benchmarks
Native audio and visual fidelity in Muse Video
Meta AI research and Superintelligence Labs
Content Seal and AI media provenance
FAQ