Krea Enhancer vs Real-ESRGAN
Krea Enhancer
krea.aiReal-ESRGAN
github.comSide by side
| Figure | Krea Enhancer | Real-ESRGAN |
|---|---|---|
| Price from | $5/mo | — |
| Free plan | Yes | — |
| Free trial | — | — |
| Open source | — | Yes |
| Web | Yes | Yes |
| Windows | — | Yes |
| macOS | — | Yes |
| Linux | — | Yes |
| iOS | Yes | — |
| API | Yes | — |
| Self-hosted | Yes | Yes |
| Plans published | 13 | 0 |
| Spec score · Image Upscaling Software | 9.3 (#1) | 8.2 (#33) |
| Batch processing | Krea advertises parallel processing of multiple image-upscaling jobs. | Yes |
| Custom training | Users can train custom LoRA styles, characters or objects from uploaded images. | The project supports finetuning on users' own data or paired data. |
| Free plan | Yes | Yes |
| Input formats | The upload page lists PNG, JPEG, WebP, HEIC, and HEIF formats. | The executable accepts JPG, PNG, and WebP images. |
| Integrations | The API site provides an MCP endpoint and setup command for Claude Code, and lists integrations with Codex, OpenClaw, Cursor and Hermes. | The README says Real-ESRGAN was integrated into Hugging Face Spaces with Gradio and links to a web demo. |
| License | Krea 2 weights use the Krea 2 Community License rather than a standard open-source license; the official inference repository itself is Apache-2.0. | The repository includes a BSD 3-Clause license that permits redistribution and use in source and binary forms subject to its conditions. |
| Purpose | Krea Enhancer sharpens details, improves clarity, and upscales images, including blurry, pixelated, or low-resolution images. | Real-ESRGAN develops practical algorithms for general image and video restoration, extending ESRGAN using training with synthetic data. |
| Upscaling method | ai | ai |
Plans
Krea Enhancer
- FreeFree
- Basic$9 a month
- Pro$35 a month
- Max$105 a month
- Business$200 a month
- FreeFree
- FreeFree
- Basic$5 a month
- Pro$21 a month
- Max$63 a month
In their own words
Krea Enhancer
Krea Enhancer uses AI to sharpen image details, improve clarity, and upscale blurry, pixelated, or low-resolution images. The workflow is to upload an image, enhance it, and download the result. It accepts PNG, JPEG, WebP, HEIC, and HEIF files and supports batch processing and API access. The Enhancer page says images can be upscaled to 8K in seconds, while paid plans list upscaling up to 22K, including paid options such as Topaz Standard. Free online use does not require signup or a credit card, but the number of free enhancements is limited; the Free plan's basic upscaling is limited to 2K resolution. The Free plan includes 100 compute units per day. Basic costs $9 a month for 5,000 units; Pro costs $35 a month for 20,000; Max costs $105 a month for 60,000; Business costs $200 a month for 80,000 and up to 50 seats. Krea lists plugins for Photoshop, After Effects, Premiere Pro, DaVinci Resolve, and Blender. Enhanced images may be used commercially.
Real-ESRGAN
Real-ESRGAN is a free, open-source project developing algorithms for image and video restoration. It offers pretrained models for general images, anime images and animation video, including AnimeVideo-v3 and smaller anime models. Image inputs include JPG, PNG and WEBP; inference can handle alpha-channel, grayscale and 16-bit images, and outputs can be JPG, PNG or WebP. Portable inference supports scale ratios of 2, 3 or 4, with 4 as the default; the Python implementation also offers arbitrary output scaling. Portable NCNN executables are available for Windows, Linux and macOS, with required binaries and models included and no CUDA or PyTorch environment needed. These executables do not include every Python feature, including outscale. Python use requires Python 3.7 or newer and PyTorch 1.7 or newer. Online inference is available through an ARC demo and Colab demos. GFPGAN is integrated for face enhancement, and the training code supports fine-tuning on user data or paired data. The project uses synthetic training data and is released under the BSD-3-Clause license.