Fabric 1.0
Fabric 1.0: weave your video ideas into reality.
A tradução deste card ainda não está disponível; parte do conteúdo aparece em inglês.
Fabric 1.0 is a powerful video generation model that creates short video clips from text prompts. It excels at producing coherent motion, consistent characters, and realistic physics, making it ideal for rapid prototyping and creative exploration. Whether you need a quick animation for a social media post or a dynamic background for a presentation, Fabric 1.0 delivers smooth, visually appealing results.
Use Fabric 1.0 when you need to bring a static idea to life—turn a description of a "sunset over a cyberpunk city" into a looping video, or visualize a product in motion. The model is particularly good at handling cinematic camera movements and maintaining style consistency across frames.
What sets Fabric 1.0 apart is its ability to understand complex scene compositions and generate videos that feel intentional, not random. It balances creativity with control, allowing you to guide the output through detailed prompts while still surprising you with novel interpretations.
- Generating short social media video clips
- Creating animated storyboards for pitches
- Producing dynamic product demonstrations
- Designing looping background visuals
- Rapidly prototyping video concepts
Objects and characters move naturally without jarring jumps or distortions.
Visual style remains stable throughout the entire clip.
Accurately interprets detailed descriptions of scenes and actions.
- Use descriptive action words like 'gliding', 'spinning', or 'flowing'.
- Specify camera movement (e.g., 'slow pan up', 'dolly zoom') for cinematic results.
- Keep prompts focused on one main subject for best coherence.
- Experiment with different lighting descriptions to set the mood.
- Generates clips up to 10 seconds long only.
- May struggle with complex multi-character interactions.
- Cannot generate audio or speech.
Aerial view of a futuristic city at sunset, flying cars between skyscrapers, warm golden light, cinematic
A robot walking through a rainy street, neon reflections on its metal body, film grain, moody
A dragon flying over a medieval castle, clouds parting, epic scale, slow motion
Frames per second; higher values give smoother motion.
A number that locks in the random elements, so you can recreate the same video later.
The number of processing steps; more steps can improve quality but take longer.
How long the video will be, in seconds (max 10).
The shape of the video, like '16:9' for widescreen or '1:1' for square.
How closely the model follows your prompt—higher values mean stricter adherence.
Os preços exibidos são o que você paga na Muitto. Sem surpresas.
How long can the generated videos be?
Fabric 1.0 produces clips up to 10 seconds in length.
Can I control the camera movement?
Yes, you can specify camera motions like pan, tilt, or zoom in your prompt.
Does the model support multiple characters?
It can handle simple interactions but may struggle with complex group scenes.