Name: GPEN API: Face Restoration, Colorization and Inpainting
Brand: magicapi
SKU: api-magicapi-blind-face-restoration-api-yangxy-gpen
Availability: InStock
Author: API Market

Shashank Agarwal

# 🧑 Face Restoration API (yangxy GPEN)

## Overview
GPEN (GAN Prior Embedded Network) is a face model for blind face restoration in the wild: it improves faces in real photos without knowing exactly how the photo was degraded. This listing is packaged by its author, yangxy.

Send an `image` and pick a `task`: `Face Restoration` (default), `Face Colorization` or `Face Inpainting`. The API returns a list of image URLs; with the default settings the list holds one image.

## Features
- **Face restoration**: improve blurry, noisy or low-quality faces in a photo, including group photos.
- **Face colorization**: add color to faces in black-and-white images.
- **Face inpainting**: fill in damaged areas of a face. With `broken_image` set to `false`, the model adds random brush strokes to simulate damage and returns the damaged and fixed images.
- **Individual faces**: `output_individual` also returns each enhanced face as its own image.

## Use Cases
1. **Old photos**: bring out faces in old or low-quality family and archive photos.
2. **Black-and-white portraits**: colorize faces in monochrome images.
3. **Damaged portraits**: repair scratches or missing areas on faces.
4. **Face crops**: get each restored face as a separate image for profiles or albums.

## Visual Examples
### Example Input
A black-and-white group photo from 1927 with the default task, `Face Restoration`:

![GPEN example input: a 1927 black-and-white group photo](https://cdn.api.market/vUvylJ5l6pthqWvDU7BwbggLDhS-noW2/-/mgxm/9e77a2c1-3448-487e-8899-1843a52a70e2/Solvay_conference_1927.png)

```json
{
  "task": "Face Restoration",
  "image": "https://cdn.api.market/vUvylJ5l6pthqWvDU7BwbggLDhS-noW2/-/mgxm/9e77a2c1-3448-487e-8899-1843a52a70e2/Solvay_conference_1927.png",
  "broken_image": true,
  "output_individual": false
}
```

### Example Output
One restored image:

![GPEN example output: the group photo with restored faces](https://blog.api.market/wp-content/uploads/2026/10/gpen-output-fnj3t910-api-market.jpg)

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `image` | string (URL) | Yes | none |  | Input image URL |
| `task` | string | No | `Face Restoration` | `Face Restoration`, `Face Colorization`, `Face Inpainting` | Task to run |
| `output_individual` | boolean | No | `false` |  | Whether to also return each enhanced face as a separate image |
| `broken_image` | boolean | No | `true` |  | For Face Inpainting: whether the input image is already damaged. When `true`, the output is the fixed image. When `false`, the model adds random brush strokes to simulate damage and the output is the damaged image plus the fixed image. |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.032 per API unit.
- **1 API unit = 1 successful request with up to 128 seconds of processing.** A request that takes longer is billed one more unit for each started block of 128 seconds (for example, 257 seconds of processing is 3 units, $0.096).
- A request uses more than one unit only when it runs past 128 seconds; the example group photo took 6.3 seconds.
- **Example cost**: the example request above took 6.3 seconds of processing, so it used 1 API unit and cost $0.032.
- **Failed or cancelled predictions are not billed.**
- **GET requests to check prediction status** are free and don't count toward API unit usage.
- **No monthly subscription fee.**

## API Documentation

### Base URL
All API requests go to: `https://prod.api.market/api/v1/magicapi/blind-face-restoration-api-yangxy-gpen`

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

### Asynchronous Processing
**IMPORTANT**: This API is asynchronous. The flow is:
1. Send a POST request to `/predictions`. A successful request returns **201 Created** with a prediction `id` and a `status` of `starting` or `processing`.
2. Poll GET `/predictions/{id}` with that ID.
3. Keep polling while `status` is `starting` or `processing`.
4. Stop when `status` is `succeeded` (read `output`), `failed` or `canceled` (read `error`).

A polling interval of 1 second and an overall timeout of 600 seconds (10 minutes) work well.

### Endpoints

#### Create a Prediction
```bash
curl -X POST "https://prod.api.market/api/v1/magicapi/blind-face-restoration-api-yangxy-gpen/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "cf4e15a70049c0119884eb2906c8ae8807af8317bea98313fefd941e414d0c91",
    "input": {
      "task": "Face Restoration",
      "image": "https://cdn.api.market/vUvylJ5l6pthqWvDU7BwbggLDhS-noW2/-/mgxm/9e77a2c1-3448-487e-8899-1843a52a70e2/Solvay_conference_1927.png",
      "broken_image": true,
      "output_individual": false
    }
  }'
```

Response (201 Created):
```json
{
  "id": "fnj3t910f1rga0d1d1gv8hf93g",
  "version": "cf4e15a70049c0119884eb2906c8ae8807af8317bea98313fefd941e414d0c91",
  "status": "starting",
  "created_at": "2026-10-23T10:49:18.373Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/blind-face-restoration-api-yangxy-gpen/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "fnj3t910f1rga0d1d1gv8hf93g",
  "status": "succeeded",
  "output": [
    "https://blog.api.market/wp-content/uploads/2026/10/gpen-output-fnj3t910-api-market.jpg"
  ],
  "error": null,
  "created_at": "2026-10-23T10:49:18.373Z",
  "started_at": "2026-10-23T10:53:10.899Z",
  "completed_at": "2026-10-23T10:53:17.207Z"
}
```

`output` is a list of image URLs. With the default settings it holds one image; `output_individual` adds the individual faces, and `Face Inpainting` with `broken_image` set to `false` adds the damaged image.

### Status Codes
- **201 Created**: prediction created (POST `/predictions`). This is the normal success response.
- **200 OK**: prediction details returned (GET `/predictions/{id}`).
- **400 Bad Request**: invalid input, for example a missing `image` or a `task` that is not one of `Face Restoration`, `Face Colorization` or `Face Inpainting`.
- **401 Unauthorized**: missing or invalid API key.
- **404 Not Found**: no prediction with that ID.
- **429 Too Many Requests**: rate limit exceeded.
- **500 Internal Server Error** or **503 Service Unavailable**: temporary server problem; retry later.

## 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.

## Error Handling
- Check the HTTP status code of every request and read the response body for the error message.
- A prediction can also end with `status` set to `failed`; the `error` field then explains why. Failed predictions are not billed.
- Retry 429, 500 and 503 responses with exponential backoff. Fix the input before retrying a 400.

## Implementation Guide
### Example Code

#### JavaScript
```javascript
const apiKey = "YOUR_API_MARKET_KEY";
const baseUrl = "https://prod.api.market/api/v1/magicapi/blind-face-restoration-api-yangxy-gpen";

async function createAndWaitForPrediction(input) {
  const response = await fetch(`${baseUrl}/predictions`, {
    method: "POST",
    headers: {
      "Content-Type": "application/json",
      "x-api-market-key": apiKey
    },
    body: JSON.stringify({ version: "cf4e15a70049c0119884eb2906c8ae8807af8317bea98313fefd941e414d0c91", input })
  });

  if (response.status !== 201) {
    throw new Error(`API error: ${response.status} ${await response.text()}`);
  }

  const prediction = await response.json();
  return pollForResults(prediction.id);
}

const delay = (ms) => new Promise((resolve) => setTimeout(resolve, ms));

async function pollForResults(predictionId) {
  const maxTimeout = 600000;
  const pollInterval = 1000;
  const startTime = Date.now();

  while (Date.now() - startTime < maxTimeout) {
    const response = await fetch(`${baseUrl}/predictions/${predictionId}`, {
      headers: { "x-api-market-key": apiKey }
    });

    if (!response.ok) {
      throw new Error(`API error: ${response.status} ${await response.text()}`);
    }

    const prediction = await response.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 delay(pollInterval);
  }

  throw new Error("Prediction timed out after 10 minutes");
}

createAndWaitForPrediction({
  "task": "Face Restoration",
  "image": "https://cdn.api.market/vUvylJ5l6pthqWvDU7BwbggLDhS-noW2/-/mgxm/9e77a2c1-3448-487e-8899-1843a52a70e2/Solvay_conference_1927.png",
  "broken_image": true,
  "output_individual": false
}).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/blind-face-restoration-api-yangxy-gpen"


def create_and_wait_for_prediction(input_data):
    headers = {"Content-Type": "application/json", "x-api-market-key": API_KEY}
    response = requests.post(
        f"{BASE_URL}/predictions",
        headers=headers,
        json={"version": "cf4e15a70049c0119884eb2906c8ae8807af8317bea98313fefd941e414d0c91", "input": input_data},
    )

    if response.status_code != 201:
        raise Exception(f"API error: {response.status_code} {response.text}")

    return poll_for_results(response.json()["id"])


def poll_for_results(prediction_id):
    max_timeout = 600
    poll_interval = 1
    headers = {"x-api-market-key": API_KEY}
    start_time = time.time()

    while time.time() - start_time < max_timeout:
        response = requests.get(f"{BASE_URL}/predictions/{prediction_id}", headers=headers)

        if response.status_code != 200:
            raise Exception(f"API error: {response.status_code} {response.text}")

        prediction = response.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(poll_interval)

    raise Exception("Prediction timed out after 10 minutes")


print(create_and_wait_for_prediction({
    "task": "Face Restoration",
    "image": "https://cdn.api.market/vUvylJ5l6pthqWvDU7BwbggLDhS-noW2/-/mgxm/9e77a2c1-3448-487e-8899-1843a52a70e2/Solvay_conference_1927.png",
    "broken_image": True,
    "output_individual": False
}))
```

## FAQ
- **Which task should I use?**
  - `Face Restoration` for blurry or low-quality faces, `Face Colorization` to add color to faces in black-and-white images, and `Face Inpainting` to fill in damaged areas of a face.
- **Does it work on group photos?**
  - Yes. The example is a group photo with many faces, returned as one restored image. Set `output_individual` to `true` to also get each face separately.
- **What does `broken_image` do?**
  - It only matters for `Face Inpainting`. Leave it `true` when your image is already damaged. Set it to `false` to have the model add random brush strokes first, which is useful for testing; you then get the damaged and the fixed image.
- **How is a request billed?**
  - $0.032 per request for up to 128 seconds of processing, plus one unit for each further started block of 128 seconds. The example took 6.3 seconds and cost $0.032.
- **Is the 201 response from POST an error?**
  - No. 201 Created means the prediction was accepted. Use the returned `id` to poll GET `/predictions/{id}` until `status` is `succeeded`, `failed` or `canceled`.
- **What if my prediction fails?**
  - Failed predictions are not billed. Read the `error` field, adjust the input and try again.

## Contact Us
If you have any questions, reach out to us through the chat on the bottom right of the page. You can also email our support team at hello@magicapi.com.


Restore blurry or damaged faces, colorize faces or inpaint missing parts with GPEN, a model for blind face restoration in the wild.