Wan 2.7
Alibaba's Wan video model: text-to-video and image-to-video with start and end frame control, up to 1080p.
Made with Wan 2.7
Sample outputs. Open in Studio to generate your own.
Where Wan 2.7 shines
Text to video
Type a prompt and generate a 2 to 15 second clip with no input image required.
Animate a still
Supply a start frame to set image-to-video in motion from your own image.
Frame to frame
Set both a start frame and an end frame to control how a shot opens and closes.
Strengths
- Handles both text-to-video and image-to-video in one model
- Image-to-video accepts optional start and end frames for first and last frame control
- Two output resolutions, 720p and 1080p
- Flexible clip length from 2 to 15 seconds
- Five aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4) for landscape, portrait, and square
Trade-offs
- Output is video only, with no still-image generation
- Priced per second, so longer clips cost proportionally more
- 1080p costs more per second than 720p
- Frame images are optional inputs, with no separate subject or reference-image input
- No audio output is offered in this configuration
At a glance
About Wan 2.7
Wan 2.7 is Alibaba's AI video model, available on HexGen for both text-to-video and image-to-video in a single workflow. Describe a scene in words to generate a clip from scratch, or feed it a still image to bring a frame to life.
For image-to-video, Wan 2.7 accepts an optional start frame and an optional end frame, giving you first and last frame control over how a shot opens and closes. Both inputs are optional, so the same model also runs as pure text-to-video when you just want to type a prompt.
Output ranges from 720p to 1080p, with clip lengths from 2 to 15 seconds and five aspect ratios covering landscape, portrait, and square. Pricing is per second of video, so you pay for the length you generate, with a higher rate at 1080p than at 720p.
Starting points
Copy, tweak, and run. Good prompts get you most of the way there.
A red kite climbing over a windswept coastal cliff at golden hour, camera slowly tilting up to follow it, 16:9.
Image-to-video: animate this portrait so the subject turns toward the window and smiles, gentle natural light, 9:16.
First frame an empty studio, last frame the same studio filled with morning light, smooth time-lapse transition, 4:3.
Pay only for what you render. 1 USD = 1,000 HGcoins. HGcoins never expire and failed runs refund automatically.
Wan 2.7 vs other models
Wan 2.7 stands out for combining text-to-video and image-to-video with start and end frame control in one model. Here is how it sits next to other video models in the HexGen catalog.
| Model | Quality | Speed | Cost | Choose it when |
|---|---|---|---|---|
Wan 2.7 This Alibaba | Great | Fast | Mid cost | Pick Wan 2.7 when you want both text-to-video and image-to-video, plus optional start and end frame control, at 720p or 1080p. |
Kuaishou | Best | Fast | Higher cost | Reach for Kling 3.0 when top-tier video quality is the priority. |
ByteDance | Great | Fastest | Lower cost | Choose Seedance 2.0 Turbo when you want faster, lower-cost video turnarounds. |
Related models
Frequently asked questions
It generates video from a text prompt (text-to-video) or from an image (image-to-video), producing clips of 2 to 15 seconds at 720p or 1080p.