Text to video
Generate video from a text prompt, with optional audio input. Supports synchronized audio and multi-shot storytelling.
Text to videoWan 2.7 AI Video Generator
Wan 2.7 is an AI video suite from Alibaba's Tongyi Lab, with four modes: text to video, image to video, reference to video and video editing. It accepts text, images, video and audio across its workflows and produces 720p–1080p video, bringing generation and editing into one creative workspace.
Open Wan 2.7Four Wan 2.7 video workflows
Image to video
Animate a first frame, control the first and last frames, or continue a source clip. Optional audio guides generation.
Image to videoReference to video
Use up to five combined image and video references, with an optional voice reference, to guide subjects and their appearance.
Reference to videoVideo editing
Edit a source video with text instructions and an optional reference image. Choose automatic audio or preserve the original audio.
Video editingCore features of the Wan 2.7 AI video generator
High-resolution output, frame control, visual references and instruction-based editing
Wan 2.7 AI video supports high-resolution output up to 1080p
Wan 2.7 supports 720p and 1080p output for detailed scenes, character action and camera movement. Choose a duration to suit the workflow: text and image generation support 2–15 seconds; reference generation and video editing support 2–10 seconds.
First and last frame control for consistent motion
Guide motion between a starting image and an ending image with Wan 2.7. First and last frame control helps produce smoother transitions and makes a shot's beginning and ending more predictable, supporting coherent multi-scene storytelling.
Nine-grid image references for complex scene composition
A nine-grid reference image supplies several viewpoints to guide scene layout and composition. Wan 2.7 uses these visual references to support more accurate framing and reduce visual drift in complex video workflows.
Subject and voice references for character consistency
Wan 2.7 accepts subject and voice references to help maintain a character's appearance, vocal identity and lip-sync continuity. These controls suit narrative scenes and character-driven AI video creation.
Instruction-based video editing for faster iteration
Use text instructions to edit an existing video, including object replacement, style changes and scene adjustments. Working from a source clip reduces the need to rebuild a concept from the beginning and supports repeated creative revisions.
Wan 2.7 output and input limits
Compare the input, duration and output resolution of the four workflows.
| Workflow | Input | Duration | Resolution |
|---|---|---|---|
| Text to video | Text · optional audio | 2–15 s | 720p / 1080p |
| Image to video | Frames or source clip · optional audio | 2–15 s | 720p / 1080p |
| Reference to video | Images / videos · optional voice | 2–10 s | 720p / 1080p |
| Video editing | Source video · optional reference image | 2–10 s | 720p / 1080p |
How to use Wan 2.7
From your prompt and reference materials to a video
- 01
Select Wan 2.7 and choose text, image, reference generation or video editing.
- 02
Enter your prompt and add the images, video or audio required by your selected workflow.
- 03
Choose 720p or 1080p, then set a duration within the selected workflow's limits.
- 04
Review the credit quote and submit. View and download the result when the task completes.
What can you create with Wan 2.7?
Explore ways to use Wan 2.7
Marketing and product videos
Create advertising concepts from text or use product images to guide a video.
Social media videos
Create short clips and use editing instructions to change a scene's background or style.
Storyboards and preproduction
Explore scenes, camera movements and multi-shot narratives, with references to guide characters.
Restoration and style transfer
Use video-editing instructions for colorization, footage restoration or a change of visual style.
Wan 2.7 FAQ
Learn about generation modes, reference materials, video settings and editing.
What is Wan 2.7?
Wan 2.7 is Alibaba Tongyi Lab’s AI video model suite. It combines text-to-video, image-to-video, reference-to-video and video editing, covering both video creation and changes to existing footage.
Which video modes does Wan 2.7 support?
Text-to-video creates a clip from a prompt; image-to-video animates a starting image. Reference-to-video uses images or videos to guide subjects and scenes, while video editing changes existing footage using text instructions.
What is the difference between text-to-video and image-to-video?
Text-to-video starts from a written description and does not require an image. Image-to-video starts from an image to guide the subject and composition; an optional final frame can also guide the transition toward a chosen ending.
Which input materials can I use?
Wan 2.7 supports text, images, video and audio, with different inputs for each mode. Image-to-video can use frame images and driving audio; reference-to-video uses subject references and optional voice audio; editing starts from a source video and can include a reference image.
Which output resolutions does Wan 2.7 support?
All four modes support 720p and 1080p video output. Choose the resolution in the workspace before submitting your task.
How long can a Wan 2.7 video be?
Text-to-video and image-to-video support 2–15 seconds. Reference-to-video and video editing support 2–10 seconds. The available duration depends on the mode you choose.
How does first-and-last-frame control work?
In image-to-video, a first-frame image sets the starting composition and an optional last-frame image specifies the ending. Your prompt guides the action between them. Wan 2.7 also supports continuing from an initial video clip.
Does Wan 2.7 support a nine-panel image reference?
Yes. Wan 2.7 can use one image containing multiple storyboard panels as a reference for characters, scenes and shots. A nine-panel reference is a single combined image, rather than nine separately uploaded reference files.
How many subject references can I use, and can I add a voice?
Reference-to-video requires at least one reference image or video and accepts up to five images and videos combined. These materials guide the subjects’ appearance. Optional voice audio can guide a character’s voice.
What can Wan 2.7 video editing change?
Describe the changes you want in text, such as replacing an object, changing a background, adjusting the visual style or colorizing black-and-white footage. You can also add a reference image to guide the edit.
Can video editing preserve the original audio?
Yes. Video editing can preserve the source video’s original audio or process audio automatically based on the editing instructions. Choose the audio option that fits your edit.
What does prompt expansion do?
Prompt expansion adds detail to your description and can help with short prompts. Enabling it also increases processing time.
Start creating with Wan 2.7
Create 720p–1080p videos with Wan 2.7. Explore text, image and reference generation, first/last frames, character and voice references, and video editing.
Open Wan 2.7Model documentation · Alibaba Cloud