# Wan v2.7 Text-to-Video

The Wan 2.7 release refines Alibaba Cloud's text-to-video pipeline with smoother motion, native audio, and multi-shot narrative control driven directly by how you write the prompt.

- **Model ID:** `alibaba/wan-v2.7-t2v`
- **Type:** video
- **Providers:** alibaba
- **Context window:** Not applicable
- **Maximum output tokens:** 0
- **Pricing:** See the provider details below
- **Canonical page:** https://vercel.com/ai-gateway/models/wan-v2.7-t2v

## Supported parameters

Detailed capability metadata has not been reported for this model.

## Example

```ts
import { experimental_generateVideo as generateVideo } from 'ai';

const result = await generateVideo({
  model: 'alibaba/wan-v2.7-t2v',
  prompt: 'A serene mountain lake at sunrise.'
});
```

## About

Wan v2.7 Text-to-Video is the text-to-video member of Alibaba Cloud's Wan 2.7 release. You provide a text prompt describing a scene, and Wan v2.7 Text-to-Video returns a finished clip of 2 to 15 seconds, with a default of 5 seconds, at 720p or 1080p. Five aspect ratio options cover landscape, portrait, square, and broadcast formats: 16:9, 9:16, 1:1, 4:3, and 3:4.

Multi-shot narrative control works directly through prompt language. Describe a sequence of events spanning multiple moments or locations, and Wan v2.7 Text-to-Video structures the clip with scene cuts and transitions rather than forcing everything into one continuous take. Audio generates natively in the same pass, and an optional audio input can drive the generation, which is useful when a soundtrack or voice line should shape the motion.

Compared to the 2.6 generation, the 2.7 line delivers smoother, more coherent motion, so subjects and camera moves hold together more reliably across the frame. That improvement compounds in multi-shot output, where consistency across cuts determines whether a generated sequence reads as one cohesive piece.

## What to consider

Pricing is per second of generated video and varies by resolution, so duration and output size are your two main cost levers. A 15-second 1080p clip sits at the top of the range; shorter 720p clips are the economical draft tier.

Run a few test prompts in the AI Gateway playground before full integration to calibrate generation times and confirm the output style matches your production bar.

## When to use

### Best For

- Production short-form video for advertising, social content, or product demos generated from a prompt alone
- Multi-scene narrative prompts where scene cuts and transitions should come from the prompt, not manual editing
- Audio-led generation where a supplied soundtrack or voice line should drive the motion of the clip
- Vertical, square, and broadcast deliverables using the five supported aspect ratios

### Consider Alternatives When

- If a real subject's appearance and voice must carry into the output, wan-v2.7-r2v handles reference-based generation
- If you want to animate an existing still image, the Wan image-to-video models are built for that workflow
- If your pipeline is already tuned to the previous generation, wan-v2.6-t2v remains available in the catalog

## Best for

- **Production Short-Form Video:** Advertising, social content, and product demos generated from a prompt alone
- **Multi-Scene Narrative Prompts:** Scene cuts and transitions structured by prompt language instead of manual editing
- **Audio-Led Generation:** Clips where a supplied soundtrack or voice line should drive the motion
- **Vertical and Broadcast Formats:** Deliverables across the five supported aspect ratios from 9:16 to 4:3

## Consider alternatives

- **Reference-Based Identity:** wan-v2.7-r2v carries a real subject's appearance and voice from reference material into the output
- **Animating an Existing Image:** The Wan image-to-video models are built for image-driven workflows
- **Previous-Generation Pipelines:** wan-v2.6-t2v remains available for workflows already tuned to the earlier release

## Frequently asked questions

### What durations and resolutions does Wan v2.7 Text-to-Video support?

Clips run 2 to 15 seconds, with a default of 5 seconds, at 720p or 1080p. Five aspect ratios are available: 16:9, 9:16, 1:1, 4:3, and 3:4.

### How does multi-shot narrative control work in Wan v2.7 Text-to-Video?

You control shot structure through prompt language. Describe a sequence spanning multiple moments or locations, and Wan v2.7 Text-to-Video introduces scene cuts and transitions to match the narrative rather than rendering one continuous take.

### Does Wan v2.7 Text-to-Video generate audio?

Yes. Audio generates natively in the same pass as the video. You can also supply an optional audio input to drive the generation when a soundtrack or voice line should shape the motion.

### How does Wan v2.7 Text-to-Video differ from the previous Wan text-to-video model?

Wan v2.7 Text-to-Video delivers smoother, more coherent motion than the 2.6 generation and supports audio input to drive generation. Resolution and aspect ratio coverage carry over, and duration is adjustable anywhere from 2 to 15 seconds.

### How do I use Wan v2.7 Text-to-Video through AI Gateway?

Call Wan v2.7 Text-to-Video with `generateVideo` from the AI SDK, passing your prompt plus optional resolution and duration settings. You can also try prompts first in the playground on this page.

### Does AI Gateway offer Zero Data Retention for Wan v2.7 Text-to-Video?

Yes, Zero Data Retention is available for this model. Zero Data Retention is offered on a per-provider basis. See https://vercel.com/docs/ai-gateway/capabilities/zdr for details.

## Links

- [Model page](https://vercel.com/ai-gateway/models/wan-v2.7-t2v)
- [AI Gateway documentation](https://vercel.com/docs/ai-gateway)
- [Provider model documentation](https://wan.video/)
- [Provider pricing](https://modelstudio.console.alibabacloud.com/ap-southeast-1/?tab=doc#/doc/?type=model&url=prices)


If you encounter incorrect information, can’t find what you need, need a capability Vercel doesn’t currently offer, or have any other feedback to report, please do it by following the instructions at https://vercel.com/feedback.md.