Wan 2.7 isn't available on SkyVid yet
Wan 2.7 is Alibaba's earlier video model for prompt-driven motion across products, environments, and stylized scenes. SkyVid does not offer Wan 2.7 video for new generations. Older jobs can still be viewed.
Opens the generator with Wan 3.0 selected. Older Wan 2.7 jobs can still be viewed in My Assets.
Alibaba
What Wan 2.7 is known for
Wan 2.7 video is known for adapting one concept across several aspect ratios. SkyVid does not run Wan 2.7 video for new clips. Wan 2.7 Image is a different, real image endpoint. The notes below describe the video model itself.
Prompt-driven motion planning
Describe subject action, environment response, camera behavior, and timing in a single compact scene brief.
One concept, several formats
Use the available ratios to adapt an idea for landscape, portrait, square, and extra-wide delivery.
Text or start-image control
Begin from a blank scene description or anchor the composition with one approved image.
Prompt notes
How people prompt Wan 2.7
These notes describe the original model. They are not a SkyVid generation workflow.
Environment choreography
A glass greenhouse stands in heavy summer rain. Camera glides between wet leaves, droplets roll across the lens edge, interior lights flicker on one row at a time, finish on a wide exterior view.
Give environmental motion a sequence rather than a vague mood.
Transformation shot
A sheet of recycled paper folds itself into a geometric desk lamp. Top-down camera slowly lowers to eye level, folds remain clean and physical, warm light turns on at the end.
State the initial material, transformation order, and final object.
Stylized image motion
Preserve the ink illustration style. The painted river begins flowing, small boats drift forward, mist separates into depth layers, and the camera eases toward the distant bridge.
Protect the source medium while animating depth and direction.
Models you can generate with
Seedance 2.5
Seedance 2.5 is SkyVid's flagship video model. It runs the real bytedance/seedance-2-5 endpoint, with audio and multimodal references, from 4 to 30 seconds.
Seedance 2.0 Mini
Use the lighter Seedance option for drafts, social clips, and motion tests where iteration speed matters more than the highest available resolution.
Grok Imagine
Explore dynamic visual ideas quickly with Grok Imagine, using text or one start image and the same SkyVid controls for format, duration, and cost.