> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.empiriolabs.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.empiriolabs.ai/_mcp/server.

# Wan 2.7

> Multimodal video model supporting T2V, I2V, video editing, and reference-to-video, with high-fidelity output from text, image, or video inputs.

![Wan 2.7](https://media.empiriolabs.ai/model-logos/wan.png)

[Alibaba Cloud](/providers/alibaba) · Video Generation

`POST /v1/videos/generations`

Multimodal video model supporting T2V, I2V, video editing, and reference-to-video, with high-fidelity output from text, image, or video inputs.

## At a glance

| Field               | Value                                         |
| ------------------- | --------------------------------------------- |
| Model id            | `wan-2-7`                                     |
| Model release date  | 2026-04-26                                    |
| Input modalities    | Text, Image, Video, Audio                     |
| Output modalities   | Video                                         |
| Context window      | -                                             |
| Weight precision    | -                                             |
| Region              | Singapore                                     |
| Features            | audio_sync, character_consistency, multi_shot |
| Native inference    | No                                            |
| New                 | No                                            |
| Supported endpoints | `POST /v1/videos/generations`                 |
| Alternate model ids | `alibaba/wan-2.7`, `wan-2.7`                  |

## Pricing

| Charge          | Spec       | Rate    |
| --------------- | ---------- | ------- |
| All Modes 720P  | per second | \$0.10  |
| All Modes 1080P | per second | \$0.150 |

## Example request

```bash
curl https://api.empiriolabs.ai/v1/videos/generations \
  -H 'Authorization: Bearer $EMPIRIOLABS_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{"model": "wan-2-7", "prompt": "sunrise over the ocean", "duration": 6}'
```

## Parameters

| Parameter         | Type    | Required | Default   | Description                                                                                                                                                                                                           |
| ----------------- | ------- | -------- | --------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| `prompt`          | string  | yes      | -         | Scene description.                                                                                                                                                                                                    |
| `mode`            | enum    | no       | `"auto"`  | t2v: text-to-video. i2v: animate the attached image. videoedit: edit the attached video. r2v: reference-to-video. · Allowed: `auto`, `t2v`, `i2v`, `videoedit`, `r2v`                                                 |
| `resolution`      | enum    | no       | `"1080p"` | Output resolution. Larger = higher fidelity but slower / more expensive. · Allowed: `720p`, `1080p`                                                                                                                   |
| `duration`        | number  | no       | `5`       | Clip length in seconds. Video edit and reference-to-video are capped at 10 seconds, and video edit never exceeds the input clip. Image-to-video with an attached audio clip matches the audio length. · Range: 2 – 15 |
| `aspect_ratio`    | enum    | no       | -         | Optional. If omitted, the model picks based on input. · Allowed: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`                                                                                                                  |
| `image`           | string  | no       | -         | Reference image URL. Required for i2v / r2v / videoedit.                                                                                                                                                              |
| `video`           | string  | no       | -         | Reference video URL. Required for videoedit / r2v.                                                                                                                                                                    |
| `negative_prompt` | string  | no       | `""`      | What to avoid.                                                                                                                                                                                                        |
| `seed`            | number  | no       | -         | Reproducibility seed.                                                                                                                                                                                                 |
| `audio_setting`   | enum    | no       | `"auto"`  | auto: generate native audio. origin: keep the audio of the source video (video edit only). · Allowed: `auto`, `origin`                                                                                                |
| `prompt_extend`   | boolean | no       | true      | Automatically expand and refine the prompt for better results.                                                                                                                                                        |

## Notes

Generation can take 30+ minutes. Modes auto-detected from attachments, or override with the `mode` parameter.

**Modes**

* T2V: no attachments
* I2V (First Frame): 1 image
* I2V (First + Last): exactly 2 images
* I2V Continuation: 1 video (2-10s) + optional last-frame image
* Video Edit: 1 video (2-10s, ≤100 MB, MP4/MOV) + up to 3 reference images
* R2V: up to 5 references combined; reference subjects in your prompt with `Video1`, `Image1`, etc.

**Optional audio**

* T2V/I2V: 2-30s
* R2V: 1-10s (used as a voice timbre sample)
* Max 15 MB, .mp3 or .wav

**Billing**

* Video Edit and R2V are billed for input + output duration combined.

**Uploaded media preprocessing**

* Reference and edit videos are normalized to provider-compatible MP4 when needed.
* Reference-video duration follows the mode limits shown above.

---

*Machine-readable schema:* `GET https://api.empiriolabs.ai/v1/models/wan-2-7`.