Name: Ultimate SD Upscale API: Tiled AI Image Upscaler
Brand: magicapi
SKU: api-magicapi-ultimate-image-upscaler-fewjative-ultimate-sd-upscale
Availability: InStock
Author: API Market

Shashank Agarwal

# 🔍 Image Upscaler API (fewjative Ultimate SD Upscale)

## Overview
Ultimate SD Upscale enlarges an image and then re-draws it tile by tile with Stable Diffusion, guided by ControlNet Tile so each tile stays close to the original. The result is a larger image with added detail.

Pick the first-pass upscaler (`upscaler`), the scale factor (`upscale_by`, default `2`), a prompt that describes the image, and how much new detail each tile may add (`denoise`). Seam fix options hide the borders between tiles.

## Features
- **Upscaling models**: `4x-UltraSharp` (default), `4x_NMKD-Siax_200k`, `RealESRGAN_x4plus` and `RealESRGAN_x4plus_anime_6B`.
- **Scale factor**: `upscale_by` (default `2`).
- **Tile re-detailing**: Stable Diffusion redraws each tile; `denoise` (default `0.2`) controls how much it may change.
- **ControlNet Tile**: `use_controlnet_tile` and `controlnet_strength` keep tiles faithful to the source.
- **Prompts**: `positive_prompt` and `negative_prompt` steer the added detail.
- **Seam fixing**: `seam_fix_mode` with `Band Pass`, `Half Tile` or `Half Tile + Intersections`.
- **Sampler control**: 20 samplers, 6 schedulers, `steps`, `cfg` and `seed`.

## Use Cases
1. **Print and large displays**: enlarge generated or photographed images for posters and banners.
2. **AI art finishing**: add detail to images from text-to-image models before publishing.
3. **Game and illustration assets**: upscale textures and artwork; `RealESRGAN_x4plus_anime_6B` suits anime art.
4. **Old or small images**: enlarge low-resolution images for reuse.

## Visual Examples
### Example Input
![Input image: a rusty mech looking down at a child with a backpack in a jungle](https://blog.api.market/wp-content/uploads/2026/10/ultimate-sd-upscale-input-api-market.jpg)

`positive_prompt`: "a dirty rust mech looking down at a kid wearing a backpack, in a jungle, low saturation, dirty, green foliage"

Other fields: `upscaler` `4x-UltraSharp`, `upscale_by` `2`, `denoise` `0.2`, `steps` `20`, `cfg` `8`, `sampler_name` `euler`, `scheduler` `normal`, `mode_type` `Linear`, `tile_width` and `tile_height` `512`, `tile_padding` `32`, `mask_blur` `8`, `use_controlnet_tile` `true`, `controlnet_strength` `1`, `force_uniform_tiles` `true`, `seam_fix_mode` `None` (seam fix settings at their defaults), empty `negative_prompt`.

### Example Output
![Ultimate SD Upscale output: the same scene at twice the size with added detail](https://blog.api.market/wp-content/uploads/2026/10/ultimate-sd-upscale-output-8ka364ss-api-market.jpg)

## Input Parameters
| Field | Type | Required | Default | Allowed values | Description |
| --- | --- | --- | --- | --- | --- |
| `image` | string (URL) | Yes |  |  | Input image to upscale |
| `positive_prompt` | string | No | `Hey! Have a nice day :D` |  | Prompt describing the image, used while re-detailing the tiles |
| `negative_prompt` | string | No | `""` |  | What to avoid while re-detailing the tiles |
| `upscaler` | string | No | `4x-UltraSharp` | `4x_NMKD-Siax_200k`, `4x-UltraSharp`, `RealESRGAN_x4plus`, `RealESRGAN_x4plus_anime_6B` | Upscaling model used to enlarge the image before the tile pass |
| `upscale_by` | number | No | `2` |  | Scale factor of the output size, for example 2 doubles the width and height |
| `use_controlnet_tile` | boolean | No | `true` |  | Use ControlNet Tile to keep each tile close to the original |
| `controlnet_strength` | number | No | `1` |  | Strength of ControlNet Tile |
| `seed` | integer | No | none |  | Sampling seed. Leave empty for a random seed |
| `steps` | integer | No | `20` |  | Sampling steps for each tile |
| `cfg` | number | No | `8` |  | Classifier-free guidance scale |
| `sampler_name` | string | No | `euler` | `euler`, `euler_ancestral`, `heun`, `dpm_2`, `dpm_2_ancestral`, `lms`, `dpm_fast`, `dpm_adaptive`, `dpmpp_2s_ancestral`, `dpmpp_sde`, `dpmpp_sde_gpu`, `dpmpp_2m`, `dpmpp_2m_sde`, `dpmpp_2m_sde_gpu`, `dpmpp_3m_sde`, `dpmpp_3m_sde_gpu`, `dpmpp`, `ddim`, `uni_pc`, `uni_pc_bh2` | Sampler |
| `scheduler` | string | No | `normal` | `normal`, `karras`, `exponential`, `sgm_uniform`, `simple`, `ddim_uniform` | Noise scheduler |
| `denoise` | number | No | `0.2` |  | Denoising strength for each tile. Low values keep the original; higher values add more new detail |
| `mode_type` | string | No | `Linear` | `Linear`, `Chess`, `None` | Tile redraw pattern: `Linear`, `Chess`, or `None` to skip the redraw |
| `tile_width` | integer | No | `512` |  | Tile width in pixels |
| `tile_height` | integer | No | `512` |  | Tile height in pixels |
| `mask_blur` | integer | No | `8` |  | Blur applied to the edges of each tile mask |
| `tile_padding` | integer | No | `32` |  | Padding around each tile in pixels |
| `seam_fix_mode` | string | No | `None` | `None`, `Band Pass`, `Half Tile`, `Half Tile + Intersections` | Extra pass that hides the seams between tiles |
| `seam_fix_denoise` | number | No | `1` |  | Denoising strength of the seam fix pass |
| `seam_fix_width` | integer | No | `64` |  | Width of the seam fix band in pixels |
| `seam_fix_mask_blur` | integer | No | `8` |  | Blur applied to the seam fix mask |
| `seam_fix_padding` | integer | No | `16` |  | Padding of the seam fix pass in pixels |
| `force_uniform_tiles` | boolean | No | `true` |  | Keep every tile the same size |

## Pricing
- **Pay As You Go (PAYG) pricing**: you only pay for what you use.
- **Price**: $0.05 per API unit.
- **1 API unit = 1 request with up to 32 seconds of processing.** A request that takes longer uses one unit per started block of 32 seconds (for example, 80 seconds of processing uses 3 units, $0.15).
- Larger inputs, higher `upscale_by`, more `steps` and seam fix passes take longer and use more units. Most requests need more than one unit.
- **Example cost**: the example request took 54.6 seconds of processing, which is 2 units, $0.10.
- **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/ultimate-image-upscaler-fewjative-ultimate-sd-upscale`

### 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/ultimate-image-upscaler-fewjative-ultimate-sd-upscale/predictions" \
  -H "Content-Type: application/json" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY" \
  -d '{
    "version": "5daf1012d946160622cd1bd45ed8f12d9675d24659276ccfe24804035f3b3ad7",
    "input": {
      "cfg": 8,
      "image": "https://blog.api.market/wp-content/uploads/2026/10/ultimate-sd-upscale-input-api-market.jpg",
      "steps": 20,
      "denoise": 0.2,
      "upscaler": "4x-UltraSharp",
      "mask_blur": 8,
      "mode_type": "Linear",
      "scheduler": "normal",
      "tile_width": 512,
      "upscale_by": 2,
      "tile_height": 512,
      "sampler_name": "euler",
      "tile_padding": 32,
      "seam_fix_mode": "None",
      "seam_fix_width": 64,
      "negative_prompt": "",
      "positive_prompt": "a dirty rust mech looking down at a kid wearing a backpack, in a jungle, low saturation, dirty, green foliage",
      "seam_fix_denoise": 1,
      "seam_fix_padding": 16,
      "seam_fix_mask_blur": 8,
      "controlnet_strength": 1,
      "force_uniform_tiles": true,
      "use_controlnet_tile": true
    }
  }'
```

Response (201 Created):
```json
{
  "id": "8ka364ss3srnc0d1d0xtj97dm4",
  "version": "5daf1012d946160622cd1bd45ed8f12d9675d24659276ccfe24804035f3b3ad7",
  "status": "starting",
  "created_at": "2026-10-23T10:07:54.328Z"
}
```

#### Check Prediction Status
```bash
curl -X GET "https://prod.api.market/api/v1/magicapi/ultimate-image-upscaler-fewjative-ultimate-sd-upscale/predictions/{id}" \
  -H "x-api-market-key: YOUR_API_MARKET_KEY"
```

Response when the prediction has finished (200 OK):
```json
{
  "id": "8ka364ss3srnc0d1d0xtj97dm4",
  "status": "succeeded",
  "output": "https://blog.api.market/wp-content/uploads/2026/10/ultimate-sd-upscale-output-8ka364ss-api-market.jpg",
  "error": null,
  "created_at": "2026-10-23T10:07:54.328Z",
  "started_at": "2026-10-23T10:08:11.077Z",
  "completed_at": "2026-10-23T10:09:05.654Z"
}
```

`output` is a single URL string pointing to the upscaled 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 an `upscaler` that is not in the list.
- **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/ultimate-image-upscaler-fewjative-ultimate-sd-upscale";

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: "5daf1012d946160622cd1bd45ed8f12d9675d24659276ccfe24804035f3b3ad7", 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({
  "cfg": 8,
  "image": "https://blog.api.market/wp-content/uploads/2026/10/ultimate-sd-upscale-input-api-market.jpg",
  "steps": 20,
  "denoise": 0.2,
  "upscaler": "4x-UltraSharp",
  "mask_blur": 8,
  "mode_type": "Linear",
  "scheduler": "normal",
  "tile_width": 512,
  "upscale_by": 2,
  "tile_height": 512,
  "sampler_name": "euler",
  "tile_padding": 32,
  "seam_fix_mode": "None",
  "seam_fix_width": 64,
  "negative_prompt": "",
  "positive_prompt": "a dirty rust mech looking down at a kid wearing a backpack, in a jungle, low saturation, dirty, green foliage",
  "seam_fix_denoise": 1,
  "seam_fix_padding": 16,
  "seam_fix_mask_blur": 8,
  "controlnet_strength": 1,
  "force_uniform_tiles": true,
  "use_controlnet_tile": true
}).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/ultimate-image-upscaler-fewjative-ultimate-sd-upscale"


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": "5daf1012d946160622cd1bd45ed8f12d9675d24659276ccfe24804035f3b3ad7", "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")


output = create_and_wait_for_prediction({
    "cfg": 8,
    "image": "https://blog.api.market/wp-content/uploads/2026/10/ultimate-sd-upscale-input-api-market.jpg",
    "steps": 20,
    "denoise": 0.2,
    "upscaler": "4x-UltraSharp",
    "mask_blur": 8,
    "mode_type": "Linear",
    "scheduler": "normal",
    "tile_width": 512,
    "upscale_by": 2,
    "tile_height": 512,
    "sampler_name": "euler",
    "tile_padding": 32,
    "seam_fix_mode": "None",
    "seam_fix_width": 64,
    "negative_prompt": "",
    "positive_prompt": "a dirty rust mech looking down at a kid wearing a backpack, in a jungle, low saturation, dirty, green foliage",
    "seam_fix_denoise": 1,
    "seam_fix_padding": 16,
    "seam_fix_mask_blur": 8,
    "controlnet_strength": 1,
    "force_uniform_tiles": True,
    "use_controlnet_tile": True
})
print(output)
```

## FAQ
- **How do I keep the result faithful to my image?**
  - Keep `denoise` low (the default `0.2`) and leave `use_controlnet_tile` on. Higher `denoise` adds more new detail but can change the content.
- **Which upscaler should I pick for anime art?**
  - `RealESRGAN_x4plus_anime_6B`. For photos and general images, the default `4x-UltraSharp` is a good start.
- **I can see lines between tiles. What do I change?**
  - Set `seam_fix_mode` to `Band Pass`, `Half Tile` or `Half Tile + Intersections`. Seam fixing adds processing time.
- **What did the example cost?**
  - It took 54.6 seconds of processing. One unit covers up to 32 seconds, so it used 2 units, $0.10.
- **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.


Enlarge an image with an upscaling model, then re-draw it tile by tile with Stable Diffusion and ControlNet Tile, with seam fixing and full sampler control.