Choose a model
Pick the model that fits the shot; the table above shows lengths, resolutions and sound.
Use case
Sky Studio is an AI video generator with six video models in one workspace: Veo 3.1 Fast, Veo 3.1 Pro and Gemini Omni 1.1 Flash from Google, Seedance 2.0 Pro from ByteDance, MiniMax H3 (Hailuo 3) and Grok Imagine from xAI. Start from a prompt, from start and end frames, or from reference images, and generate clips of up to 15 seconds and up to 4K, with sound on the models that support it.
Generate a video| Model | Resolution | Clip length | Sound | Start from | Best for |
|---|---|---|---|---|---|
| Veo 3.1 FastGoogle | 1080pβ4K | 8 s | Yes | Text, start and end frames | Realistic clips with native sound, quickly |
| Veo 3.1 ProGoogle | 1080pβ4K | 8 s | Yes | Text, start and end frames | The highest-quality Veo clips, with native sound |
| Seedance 2.0 ProByteDance | 480pβ1080p | 4β15 s | Optional | Text, start and end frames, reference images, video and audio | Longer shots guided by several references |
| MiniMax H3 (Hailuo 3)MiniMax | 480pβ4K | 5β15 s | Yes | Text, start and end frames, reference images | Lively motion; up to 4K from reference images |
| Grok ImaginexAI | 480pβ720p | 1β15 s | Yes | Reference images | Keeping a character consistent from reference images |
| Gemini Omni 1.1 FlashGoogle | 360pβ4K | 3β10 s | Yes | Text, start and end frames, reference images, extend a clip | Quick drafts at low resolution, and extending clips |
Prompt: βA red paper boat drifts down a rain-soaked city street at dusk, neon signs reflecting in the rippling water, slow tracking shot at street level, cinematic, gentle sound of rainβ
Pick the model that fits the shot; the table above shows lengths, resolutions and sound.
Write a prompt, add a start and end frame, or add reference images of your character or product.
The options follow the model you chose, and the app shows the credit cost before you generate.
Videos render in the background and land in your workspace when they are ready.
Continue in the Cinematic canvas to plan scenes, or download the clip.
It depends on the shot. Veo 3.1 is strong at realism and native sound, Seedance 2.0 Pro at longer shots guided by several references, Grok Imagine at keeping a character consistent from reference images, MiniMax H3 at lively motion, and Gemini Omni 1.1 Flash at quick drafts. In Sky Studio you can try the same idea on several models and keep the best take.
One clip runs up to 15 seconds with Seedance 2.0 Pro, MiniMax H3 and Grok Imagine, 10 seconds with Gemini Omni 1.1 Flash and 8 seconds with Veo 3.1. Longer films are built from several clips in the Cinematic canvas.
Yes. Most models take a start frame, and several also take an end frame or reference images, so a product photo or a character design can become the first shot.
Yes, on most models: Veo 3.1, Gemini Omni 1.1 Flash, MiniMax H3 and Grok Imagine return clips with sound, and on Seedance you turn sound on or off. The table above shows it for every model.
No. Sky Studio is pay as you go: you buy credits when you need them, and they never expire. The app shows the cost of each video before you generate.
Generate images with Nano Banana 2 and Pro, GPT Image 2 and Seedream 5.0 in one place. Compare the models, see samples and pick the right one.
Create product studio shots, solid backgrounds, flat lays, and multi-angle e-commerce visuals with Sky Studio.
Remove image backgrounds with Sky Studio Background Remover, running in the browser without uploading the cutout step.
Add draggable local light layers to product photos and create studio-style lighting with Sky Studio Product Relight.
Upload customer room photos, analyze the space, and compose products into real room scenes.
Generate 10 to 300 second music tracks in Sky Studio with simple mode, custom mode, lyrics, instrumentals, and duration control.
Make a four-view character turnaround sheet (front, side, back and close-up) from text or a reference image, then reuse the character in images and video.
Transform ordinary visuals into cinematic-looking concepts for shorts, ads, and creative video style exploration.