Name: Natural Language Video Editing API (Wan 2.7 VideoEdit)
Brand: magicapi
SKU: api-magicapi-natural-language-video-editing-wan-video-wan-2-7-videoedit
Availability: InStock
Author: API Market

Shashank Agarwal

# ✂️ Video Editing API with Natural Language Instructions (Wan 2.7 VideoEdit)

## Overview
The Video Editing API runs Wan 2.7 VideoEdit, a model from Alibaba's Wan team that edits videos from natural language instructions. You send a short video (mp4 or mov, 2 to 10 seconds) and a prompt with editing instructions or a style transfer description, and the model returns the edited video.

You can also pass a reference image for reference-based editing, choose 720p or 1080p output, set the aspect ratio, shorten the output, and decide whether the original audio is kept.

## Features
- Edits an input video (`video`, mp4 or mov, 2 to 10 seconds) from a text instruction (`prompt`).
- Optional `reference_image` (jpg, png, bmp or webp) for reference-based editing.
- Output resolution `720p` or `1080p` (default).
- `aspect_ratio`: `auto` (default, uses the input video's aspect ratio), `16:9`, `9:16`, `1:1`, `4:3` or `3:4`.
- Optional `duration` (2 to 10 seconds) to truncate the output; if not set, the output matches the input video's duration.
- `audio_setting`: `auto` (default, the model decides whether to regenerate audio) or `origin` (keep the original audio).
- Optional `seed` (0 to 2147483647) for reproducible generation.

## Use Cases
1. **Restyling footage**  
   Change the look of a clip, for example the season, time of day or visual style, with a one-line instruction.

2. **Content variations**  
   Produce several edited versions of the same clip for different campaigns or platforms.

3. **Reference-guided edits**  
   Use a reference image to steer how the edited video should look.

4. **Reformatting**  
   Output the edit in a different aspect ratio such as 9:16 or 1:1 for social platforms.

## Visual Examples
This is a real request sent through api.market.

**Input video** (a woman in a yellow raincoat on a rainy street at night under a pink neon sign): [Watch the input video](https://blog.api.market/wp-content/uploads/2026/10/hailuo-2.3-fast-output-api-market.mp4)

**Prompt:** "Turn the scene into a snowy winter night"

**Settings:** `resolution: "720p"`. All other fields used their defaults.

**Output video** (the same scene with falling snow and a snow-covered street): [Watch the example video](https://blog.api.market/wp-content/uploads/2026/10/wan-2.7-videoedit-output-api-market.mp4)

## Pricing
- Price: $0.125 per API unit.
- 1 API unit = 1 second of output video, rounded up to a whole second.
- The output length matches your input video unless you set `duration` to truncate it.
- Worked example: a 5 second output costs 5 units ($0.625); a 10 second output costs 10 units ($1.25).
- The price per second is the same for `720p` and `1080p`. `aspect_ratio`, `audio_setting`, `reference_image` and `seed` do not change the price.
- Failed or canceled predictions are not billed.
- GET requests to check prediction status are free and don't count toward API unit usage.
- No monthly subscription fee. Pay As You Go (PAYG): you only pay for what you use.

## API Documentation

### Base URL
All API requests go to: `https://prod.api.market/api/v1/magicapi/natural-language-video-editing-wan-video-wan-2-7-videoedit`

### Authentication
Send your api.market API key in the `x-api-market-key` header:
```
x-api-market-key: YOUR_API_MARKET_KEY
```

### Asynchronous Processing
This API is asynchronous:
1. POST `/predictions` with a `version` and an `input` object. A successful request returns **201 Created** with a prediction `id` and a `status` of `starting` or `processing`.
2. Poll GET `/predictions/{id}` (for example once per second) while the status is `starting` or `processing`.
3. Stop when the status is `succeeded` (the `output` field holds the video URL), `failed` or `canceled` (the `error` field explains why).

Video jobs can take several minutes, so allow a generous client timeout (the examples below use 10 minutes).

### Endpoints

#### Create a Prediction
```bash
curl -X POST "https://prod.api.market/api/v1/magicapi/natural-language-video-editing-wan-video-wan-2-7-videoedit/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
  "version": "e1825a94300a0a887079a54c2e27b26a6e03184a031f03d989829ed5ec9145b3",
  "input": {
    "video": "https://blog.api.market/wp-content/uploads/2026/10/hailuo-2.3-fast-output-api-market.mp4",
    "prompt": "Turn the scene into a snowy winter night",
    "resolution": "720p"
  }
}'
```

Response (201 Created):
```json
{
  "id": "ufawqhfynnddngldkgtslldrkq",
  "version": "e1825a94300a0a887079a54c2e27b26a6e03184a031f03d989829ed5ec9145b3",
  "status": "starting",
  "created_at": "2026-10-13T10:37:01.920Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/natural-language-video-editing-wan-video-wan-2-7-videoedit/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when finished (200 OK):
```json
{
  "id": "ufawqhfynnddngldkgtslldrkq",
  "status": "succeeded",
  "output": "https://blog.api.market/wp-content/uploads/2026/10/wan-2.7-videoedit-output-api-market.mp4",
  "created_at": "2026-10-13T10:37:01.920Z",
  "started_at": "2026-10-13T10:37:03.112Z",
  "completed_at": "2026-10-13T10:39:12.480Z"
}
```

### Status Codes
- 200 OK: prediction status returned.
- 201 Created: prediction created; poll it for the result. This is a success response.
- 400 Bad Request: invalid or missing input fields.
- 401 Unauthorized: missing or invalid API key.
- 404 Not Found: no prediction with that id.
- 429 Too Many Requests: rate limit exceeded; wait and retry.
- 500 Internal Server Error or 503 Service Unavailable: temporary server problem; retry with backoff.

### Error Handling
- Treat both 200 and 201 as success on the create call.
- A prediction can be accepted (201) and still end with `status: "failed"`. Read the `error` field in that case; failed predictions are not billed.
- Retry 429 and 5xx responses with exponential backoff. Do not retry 400 or 401 without fixing the request or key.

## Rate Limits

All plans allow 10 requests per second. Requests above that return `429 Too Many Requests`; back off before retrying and honour any `Retry-After` header. Polling a prediction with GET is free but still counts toward the per-second limit, so poll every 1 to 5 seconds rather than in a tight loop.

## Implementation Guide

### JavaScript
```javascript
const apiKey = "YOUR_API_MARKET_KEY";
const baseUrl = "https://prod.api.market/api/v1/magicapi/natural-language-video-editing-wan-video-wan-2-7-videoedit";

async function runPrediction(input) {
  const create = await fetch(`${baseUrl}/predictions`, {
    method: "POST",
    headers: { "Content-Type": "application/json", "x-api-market-key": apiKey },
    body: JSON.stringify({ version: "e1825a94300a0a887079a54c2e27b26a6e03184a031f03d989829ed5ec9145b3", input })
  });
  if (create.status !== 200 && create.status !== 201) {
    throw new Error(`API error: ${create.status} ${await create.text()}`);
  }
  const { id } = await create.json();

  const deadline = Date.now() + 600000; // 10 minutes
  while (Date.now() < deadline) {
    const res = await fetch(`${baseUrl}/predictions/${id}`, { headers: { "x-api-market-key": apiKey } });
    if (!res.ok) throw new Error(`API error: ${res.status} ${await res.text()}`);
    const prediction = await res.json();
    if (prediction.status === "succeeded") return prediction.output;
    if (prediction.status === "failed" || prediction.status === "canceled") {
      throw new Error(`Prediction ${prediction.status}: ${prediction.error || "unknown error"}`);
    }
    await new Promise((r) => setTimeout(r, 1000));
  }
  throw new Error("Prediction timed out after 10 minutes");
}

runPrediction({
  "video": "https://blog.api.market/wp-content/uploads/2026/10/hailuo-2.3-fast-output-api-market.mp4",
  "prompt": "Turn the scene into a snowy winter night",
  "resolution": "720p"
}).then((output) => console.log(output));
```

### Python
```python
import time
import requests

API_KEY = "YOUR_API_MARKET_KEY"
BASE_URL = "https://prod.api.market/api/v1/magicapi/natural-language-video-editing-wan-video-wan-2-7-videoedit"

def run_prediction(input_data):
    headers = {"Content-Type": "application/json", "x-api-market-key": API_KEY}
    create = requests.post(
        f"{BASE_URL}/predictions",
        headers=headers,
        json={"version": "e1825a94300a0a887079a54c2e27b26a6e03184a031f03d989829ed5ec9145b3", "input": input_data},
    )
    if create.status_code not in (200, 201):
        raise Exception(f"API error: {create.status_code} {create.text}")
    prediction_id = create.json()["id"]

    deadline = time.time() + 600  # 10 minutes
    while time.time() < deadline:
        res = requests.get(f"{BASE_URL}/predictions/{prediction_id}", headers={"x-api-market-key": API_KEY})
        if res.status_code != 200:
            raise Exception(f"API error: {res.status_code} {res.text}")
        prediction = res.json()
        if prediction["status"] == "succeeded":
            return prediction["output"]
        if prediction["status"] in ("failed", "canceled"):
            raise Exception(f"Prediction {prediction['status']}: {prediction.get('error') or 'unknown error'}")
        time.sleep(1)
    raise Exception("Prediction timed out after 10 minutes")

print(run_prediction({
  "video": "https://blog.api.market/wp-content/uploads/2026/10/hailuo-2.3-fast-output-api-market.mp4",
  "prompt": "Turn the scene into a snowy winter night",
  "resolution": "720p"
}))
```

### cURL
```bash
# Step 1: create the prediction (returns 201 Created with an id)
curl -X POST "https://prod.api.market/api/v1/magicapi/natural-language-video-editing-wan-video-wan-2-7-videoedit/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
  "version": "e1825a94300a0a887079a54c2e27b26a6e03184a031f03d989829ed5ec9145b3",
  "input": {
    "video": "https://blog.api.market/wp-content/uploads/2026/10/hailuo-2.3-fast-output-api-market.mp4",
    "prompt": "Turn the scene into a snowy winter night",
    "resolution": "720p"
  }
}'

# Step 2: poll until status is succeeded, failed or canceled (replace PREDICTION_ID)
curl -X GET "https://prod.api.market/api/v1/magicapi/natural-language-video-editing-wan-video-wan-2-7-videoedit/predictions/PREDICTION_ID" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

## Input Parameters
| Parameter | Type | Required | Default | Allowed values | Description |
|-----------|------|----------|---------|----------------|-------------|
| `video` | string (URL) | Yes | none | - | Input video to edit (mp4/mov, 2-10s) |
| `prompt` | string | Yes | none | - | Editing instructions or style transfer description |
| `reference_image` | string (URL) | No | none | - | Reference image for reference-based editing (jpg/png/bmp/webp) |
| `resolution` | string | No | `"1080p"` | `"720p"`, `"1080p"` | Output video resolution |
| `aspect_ratio` | string | No | `"auto"` | `"auto"`, `"16:9"`, `"9:16"`, `"1:1"`, `"4:3"`, `"3:4"` | Aspect ratio of the output video. 'auto' uses the input video's aspect ratio. |
| `duration` | integer | No | none | 2 to 10 | Duration of the output video in seconds. If not set, matches input video duration. Use this to truncate. |
| `audio_setting` | string | No | `"auto"` | `"auto"`, `"origin"` | Audio behavior for edited video. 'auto': model decides whether to regenerate audio. 'origin': keep the original audio. |
| `seed` | integer | No | none | - | Random seed for reproducible generation. Range: 0-2147483647 |

## Output
A single URL (string) pointing to the edited MP4 video.

## FAQ
1. **What kind of video can I send?**  
   The schema accepts mp4 or mov files between 2 and 10 seconds long, passed as a URL in `video`.

2. **Can I keep the original sound?**  
   Yes. Set `audio_setting` to `origin` to keep the original audio. With `auto` (the default) the model decides whether to regenerate audio.

3. **Can I make the output shorter than the input?**  
   Yes. Set `duration` (2 to 10 seconds). You are billed for the length of the output video.

4. **Does 1080p cost more than 720p?**  
   No. Both are billed at $0.125 per second of output video.

## Contact Us

If you have any questions or need assistance, you can reach out to us through the embedded chat on the bottom right corner of the page or send an email to hello@magicapi.com.


Edit videos with natural language instructions using Alibaba Wan 2.7 VideoEdit. 720p or 1080p output, priced per second of video.