Text to video
Build a new scene from a chronological prompt covering subject, action, camera, lighting, shot changes, and pacing.
Alibaba video model · Available on Nano Banana Pro
Wan 2.7 AI Video Generator
Create a short video from a written scene, animate one approved still, or guide a continuous transition with first and last frames.
01
LIVE Create a video with Wan 2.7
Choose text, one image, or first and last frames, then set duration, resolution, and aspect ratio before reviewing the platform point estimate.
Results
Generated content will appear here
02 / Overview
Wan 2.7 at a glance
Facts reviewed:
- Provider
- Alibaba Cloud Model Studio
- Platform routes
- wan2.7-t2v / wan2.7-i2v
- Platform inputs
- Text and frame images
- Platform output
- MP4 video
- Platform duration
- 2–15 seconds
- Platform resolution
- 720p or 1080p
Model brief
What is Wan 2.7?
Wan 2.7 is a video generation model family available through Alibaba Cloud Model Studio. Official documentation lists text-to-video with multi-shot narrative and audio synchronization, plus image-to-video for first-frame generation, first-and-last-frame generation, video continuation, and audio-driven workflows. The documented output is 720p or 1080p, 2–15 seconds, 30 fps MP4 for the core text and image routes.
Nano Banana Pro currently exposes a focused subset: text-to-video, one-image animation, and first-and-last-frame generation. This workspace does not accept custom audio or video reference files and does not expose Wan’s separate reference-to-video or video-editing routes. Its aspect ratios, point estimates, and submission requirements are platform settings.
For text-to-video, treat the prompt like a compact edit sheet. Establish the main subject and visual grammar once, then give each shot a distinct purpose: orientation, action, evidence, reaction, or reveal. Repeat only the continuity constraints that matter across cuts—identity, wardrobe, prop geometry, screen direction, weather, and color—so the model has room to vary composition without losing the story.
For image-to-video, separate what must remain fixed from what may move. Preserve identity, object proportions, architecture, material colors, and camera axis; then specify motion amplitude, timing, camera path, wind, reflections, shadows, or focus changes. A narrow motion brief usually protects a prepared key visual better than asking every element to animate at once.
Before delivery, review the generated clip at normal speed and frame by frame. Check the first and last two seconds, transitions between shot sizes, hands and contact points, edge stability, product geometry, signage, repeated background figures, exposure jumps, and unintended cuts. Keep the prompt, source permissions, settings, point estimate, version, reviewer, and intended channel with the approved output.
03 / Capabilities 01 02 03 04 05 06
Key Wan 2.7 capabilities
Choose the simplest input route that gives the shot enough structure, while keeping broader official capabilities separate from this integration.
Image to video
Animate one permitted image while preserving its important subject, style, composition, and spatial relationships.
First and last frames
Upload two compatible images to define the starting frame and visual destination of a continuous transition.
Multi-shot narrative
Official documentation highlights connected multi-shot storytelling for the Wan 2.7 text-to-video route.
Audio synchronization
Wan 2.7 officially supports audio-enabled workflows, although this page does not currently expose audio input or an audio toggle.
Two production resolutions
Choose 720p for lower platform cost or 1080p when the concept needs more output detail.
04 / Scene ideas
Popular Wan 2.7 use cases
Choose the simplest input route that gives the shot enough structure, while keeping broader official capabilities separate from this integration.
Multi-shot concept films
Draft a short sequence with connected locations, shot sizes, actions, and a consistent principal subject.
Interior and architecture motion
Animate approved visualizations with controlled camera movement, daylight changes, and restrained environmental motion.
Product and campaign clips
Turn an approved key visual into a short horizontal, square, or vertical motion concept.
Time and environment transitions
Use prepared first and last frames to explore seasons, lighting changes, transformations, and reveals.
Vertical product and retail stories
Design mobile-first reveals with controlled material movement, negative space, stable product geometry, and a deliberate final hold.
Public-space and event visualization
Explore time-of-day, lighting, crowd density, and environmental changes between compatible architectural keyframes.
05 / Prompt lab
Wan 2.7 prompt examples
Each example loads the matching platform mode so you can revise it directly in the live workspace.
SCENE 01text to video
Connected travel story
A multi-shot text brief with a consistent subject, location progression, and clear transitions.
View prompt
A quiet cinematic travel sequence following the same woman in a rust-red coat through a coastal town at sunrise. Shot 1: wide view as she walks down stone steps toward the harbor. Shot 2: medium tracking shot beside colorful fishing boats while gulls cross the background. Shot 3: close-up as she unfolds an old postcard and looks toward the lighthouse. Natural transitions, consistent face and clothing, soft sea haze, realistic movement, no captions or logos.
SCENE 02image to video
Architectural still animation
Turn an approved interior image into a restrained camera and lighting study.
View prompt
Preserve the room layout, furniture design, materials, window placement, and perspective from the uploaded image. Add a slow lateral camera move, sunlight gradually traveling across the wooden floor, subtle curtain movement, and a small shift in plant shadows. Photorealistic, stable geometry, no new objects, no text.
SCENE 03start end to video
Seasonal first-to-last transition
Connect matching compositions while controlling the environmental transformation between them.
View prompt
Create one continuous transition from the uploaded summer opening frame to the uploaded winter ending frame. Keep the building, camera position, and main tree aligned. Leaves turn gold and drift away, the sky cools, light snow begins, and the ground gradually becomes white. No cuts, smooth time progression, stable architecture, physically coherent wind and snowfall.
SCENE 04text to video
Vertical product reveal
Build a mobile-first product story with a clear opening hook, material motion, and a stable final hold.
View prompt
9:16 vertical product film for an unbranded amber glass perfume bottle. 0–3s: extreme close-up as a narrow light streak travels across the glass and condensation beads. 4–8s: camera pulls back while the bottle rises slowly through low white mist. 9–12s: orbit thirty degrees and settle on a centered hero frame with clean negative space above. Preserve bottle shape, cap alignment, liquid level, and amber color; no labels, text, hands, or extra objects.
SCENE 05start end to video
Day-to-night public space
Use compatible first and last frames to test crowd continuity, lighting progression, and a deliberate camera path.
View prompt
Connect the uploaded daytime plaza frame to the uploaded nighttime festival frame in one continuous forward glide. Keep building lines, fountain position, paving pattern, and camera height aligned. Pedestrians gradually increase without duplication, shop lights turn on in sequence, the sky shifts naturally to blue hour, and warm lanterns appear around the fountain. No cuts, no warped architecture, stable travel direction.
06 / Workflow
How to use Wan 2.7
- 01
Choose the source
Start from text, upload one approved image, or provide two visually compatible first and last frames.
- 02
Write events in order
Describe each shot or movement chronologically, including subject continuity and the intended ending.
- 03
Set output controls
Choose duration, 720p or 1080p, and an aspect ratio that matches the intended publishing surface.
- 04
Review the estimate
Inspect the current Nano Banana Pro point estimate and all selected settings before submitting the task.
Limits, pricing, and responsible use
Wan 2.7 can still produce identity drift, inconsistent objects, unstable hands, accidental cuts, or imperfect transitions. Review every frame before publishing, presenting, or delivering generated work.
Official Wan 2.7 documentation includes custom audio, video continuation, reference-to-video, and video editing capabilities that are not all available here. Nano Banana Pro currently accepts text and up to two frame images on this page. The 2–15 second range and 720p/1080p options align with core official routes; aspect ratios and point estimates describe this platform integration.
Only upload content you are authorized to use. Avoid deceptive impersonation, privacy violations, intellectual-property infringement, or presenting generated footage as documentary evidence. Preserve applicable provenance and disclosure information.
Sources and fact review
Model capabilities and regional availability can change. Official Alibaba Cloud documentation and Nano Banana Pro platform settings are listed separately below.
07 / FAQ
Wan 2.7 FAQ
Can I use Wan 2.7 online on this page?
Yes. This page opens the Nano Banana Pro video workspace with Wan 2.7 selected by default, while keeping the model selector available. An account and sufficient platform points are required to submit a task.
Which Wan 2.7 routes does this page use?
Text prompts use the platform’s wan2.7-t2v route. One-image and first-and-last-frame workflows use its wan2.7-i2v route.
Can Wan 2.7 generate video from an image?
Yes. Upload one image you have permission to use, then describe camera movement, subject motion, environmental change, pacing, and the ending.
Does Wan 2.7 support first and last frames?
Yes. Alibaba Cloud documents first-and-last-frame generation for Wan 2.7 image-to-video. Upload two compatible frames and describe the continuous transition.
Can I upload custom audio on this page?
No. Wan 2.7 officially supports audio-enabled workflows, but this Nano Banana Pro workspace currently exposes only text and frame-image inputs.
Does this page support reference-to-video or video editing?
No. Those are separate official Wan 2.7 routes. This page currently focuses on text-to-video, one-image animation, and first-and-last-frame generation.
What duration and resolution are available here?
Nano Banana Pro currently offers integer durations from 2 to 15 seconds and either 720p or 1080p output.
Is Nano Banana Pro an official Alibaba product?
No. Nano Banana Pro is an independent multi-model platform. Wan and Alibaba names and trademarks belong to their respective owners.
How should I write a multi-shot Wan 2.7 prompt?
Start with one continuity block for subject, wardrobe, color, weather, and screen direction. Then write each shot in chronological order with its framing, action, camera movement, and intended transition. Give every shot a different narrative job.
How can I protect an uploaded image while adding motion?
List the fixed elements first—identity, proportions, layout, materials, perspective, and camera axis—then request only a few controlled changes. Avoid animating every object, and inspect edges and contact points frame by frame.
Related video models and tools
Render your next scene
Open Wan 2.7 workspaceCreate your first Wan 2.7 video
Return to the live workspace, choose the right source mode, and confirm resolution, duration, and points before generating.