> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/xinntao/Real-ESRGAN/llms.txt
> Use this file to discover all available pages before exploring further.

# Basic Inference

> Learn how to use Real-ESRGAN for image super-resolution with the inference_realesrgan.py script

## Overview

The `inference_realesrgan.py` script is the main tool for performing image super-resolution with Real-ESRGAN. It supports various models, configurations, and advanced features like face enhancement and tile processing.

## Quick Start

<Steps>
  <Step title="Basic Usage">
    Run Real-ESRGAN with default settings:

    ```bash theme={null}
    python inference_realesrgan.py -n RealESRGAN_x4plus -i inputs
    ```

    This will process all images in the `inputs` folder using the `RealESRGAN_x4plus` model and save results to the `results` folder.
  </Step>

  <Step title="Custom Output">
    Specify input and output paths:

    ```bash theme={null}
    python inference_realesrgan.py -n RealESRGAN_x4plus -i input.jpg -o output_folder
    ```
  </Step>

  <Step title="View Results">
    Enhanced images will be saved with the `_out` suffix by default (e.g., `image_out.png`).
  </Step>
</Steps>

## Command-Line Arguments

### Required Arguments

<ParamField path="-i, --input" type="string" default="inputs">
  Input image file or folder containing images to process
</ParamField>

<ParamField path="-n, --model_name" type="string" default="RealESRGAN_x4plus">
  Model to use for super-resolution. Available models:

  * `RealESRGAN_x4plus` - General purpose 4x upscaling
  * `RealESRNet_x4plus` - ESRGAN model without GAN
  * `RealESRGAN_x4plus_anime_6B` - Optimized for anime images
  * `RealESRGAN_x2plus` - 2x upscaling model
  * `realesr-animevideov3` - Anime video model (XS size)
  * `realesr-general-x4v3` - General purpose with denoise control
</ParamField>

### Output Options

<ParamField path="-o, --output" type="string" default="results">
  Output folder for enhanced images
</ParamField>

<ParamField path="-s, --outscale" type="float" default="4">
  Final upsampling scale of the image. Can be any value (e.g., 3.5).

  The model performs its native upscaling, then resizes to the target scale using LANCZOS4 interpolation.
</ParamField>

<ParamField path="--suffix" type="string" default="out">
  Suffix added to restored images (e.g., `image_out.png`). Set to empty string for no suffix.
</ParamField>

<ParamField path="--ext" type="string" default="auto">
  Output image format: `auto`, `jpg`, or `png`

  * `auto`: Uses the same extension as input
  * RGBA images are always saved as PNG
</ParamField>

### Performance Options

<ParamField path="-t, --tile" type="integer" default="0">
  Tile size for processing large images. Set to 0 to disable tiling.

  <Warning>
    Use tiling if you encounter CUDA out of memory errors. Recommended values: 256, 400, or 512.
  </Warning>

  ```bash theme={null}
  python inference_realesrgan.py -n RealESRGAN_x4plus -i input.jpg --tile 400
  ```
</ParamField>

<ParamField path="--tile_pad" type="integer" default="10">
  Padding size for each tile to reduce edge artifacts
</ParamField>

<ParamField path="--pre_pad" type="integer" default="0">
  Pre-padding size at each border of the image
</ParamField>

<ParamField path="--fp32" type="flag">
  Use FP32 (full precision) instead of FP16 (half precision) during inference

  Increases memory usage but may improve quality slightly. Default is FP16.
</ParamField>

<ParamField path="-g, --gpu-id" type="integer" default="None">
  GPU device to use (e.g., 0, 1, 2 for multi-GPU systems)
</ParamField>

### Advanced Options

<ParamField path="--face_enhance" type="flag">
  Enable GFPGAN face enhancement integration

  Automatically enhances faces in the image using GFPGAN v1.3

  ```bash theme={null}
  python inference_realesrgan.py -n RealESRGAN_x4plus -i inputs --face_enhance
  ```
</ParamField>

<ParamField path="-dn, --denoise_strength" type="float" default="0.5">
  Denoise strength for the `realesr-general-x4v3` model only

  * `0`: Weak denoise (keep noise)
  * `1`: Strong denoise ability
  * Values between 0 and 1 interpolate between the two extremes
</ParamField>

<ParamField path="--model_path" type="string" default="None">
  Custom model path. Usually not needed as models are auto-downloaded to the `weights` folder.
</ParamField>

<ParamField path="--alpha_upsampler" type="string" default="realesrgan">
  Upsampler for alpha channels in RGBA images: `realesrgan` or `bicubic`
</ParamField>

## Common Usage Examples

### Example 1: Basic Upscaling

```bash theme={null}
python inference_realesrgan.py -n RealESRGAN_x4plus -i input.jpg
```

### Example 2: Custom Scale with Face Enhancement

```bash theme={null}
python inference_realesrgan.py -n RealESRGAN_x4plus -i input.jpg --outscale 3.5 --face_enhance
```

### Example 3: Batch Processing with Tiling

```bash theme={null}
python inference_realesrgan.py -n RealESRGAN_x4plus -i inputs_folder -o results --tile 400 --tile_pad 10
```

### Example 4: Denoise Control

```bash theme={null}
python inference_realesrgan.py -n realesr-general-x4v3 -i input.jpg --denoise_strength 0.8
```

## Model Downloads

Models are automatically downloaded on first use. Manual download URLs:

<Tabs>
  <Tab title="RealESRGAN_x4plus">
    ```bash theme={null}
    wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth -P weights
    ```
  </Tab>

  <Tab title="RealESRNet_x4plus">
    ```bash theme={null}
    wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.1/RealESRNet_x4plus.pth -P weights
    ```
  </Tab>

  <Tab title="RealESRGAN_x2plus">
    ```bash theme={null}
    wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.1/RealESRGAN_x2plus.pth -P weights
    ```
  </Tab>

  <Tab title="realesr-general-x4v3">
    ```bash theme={null}
    wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-x4v3.pth -P weights
    wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.5.0/realesr-general-wdn-x4v3.pth -P weights
    ```
  </Tab>
</Tabs>

## Error Handling

<Warning>
  **CUDA Out of Memory Error**

  If you encounter this error:

  ```
  RuntimeError: CUDA out of memory
  ```

  Try the following solutions:

  1. Use the `--tile` option with a smaller value (e.g., 256 or 400)
  2. Enable `--fp32` if using half precision
  3. Use a smaller `--outscale` value
  4. Process images one at a time instead of batch processing
</Warning>

## Supported Image Formats

The inference script supports:

* **Color spaces**: RGB, RGBA (with alpha channel), Grayscale
* **Bit depths**: 8-bit, 16-bit images
* **Formats**: JPG, PNG, WebP, and other common formats

## Next Steps

<CardGroup cols={2}>
  <Card title="General Images" icon="image" href="/guides/general-images">
    Learn model selection for real-world photos
  </Card>

  <Card title="Anime Images" icon="sparkles" href="/guides/anime-images">
    Optimize for anime and illustration content
  </Card>

  <Card title="Face Enhancement" icon="user" href="/guides/face-enhancement">
    Integrate GFPGAN for better face restoration
  </Card>

  <Card title="Video Processing" icon="video" href="/guides/anime-videos">
    Process video files frame by frame
  </Card>
</CardGroup>