Krea Enhancer
8.4#2 in image upscaling softwareFree planFree planiOSWebAPISelf-hosted
VisitKrea Enhancer vs Real-ESRGAN
Real-ESRGAN
7.6#18 in image upscaling softwareWebLinuxmacOSSelf-hosted
VisitWhat both publish
| Fact | Krea Enhancer | Real-ESRGAN |
|---|---|---|
| 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. |
Krea Enhancer plans
- FreeFree
- Basic$5 a month
- Pro$21 a month
- Max$63 a month
- Business$160 a month
- Maker self-hosted model license$1,000 a month
Real-ESRGAN plans
No prices published.
Pick Krea Enhancer if…
Krea Enhancer suits people restoring or enlarging portraits, landscapes, product images, family photos, and screenshots. It also fits developers and creative teams needing batch processing, API access, editor plugins, or commercial image use.
Pick Real-ESRGAN if…
Real-ESRGAN may suit developers and users seeking image or video restoration models, including tools for general images and anime. Its portable executables offer a route that does not require CUDA or PyTorch, while Python users can access arbitrary output scaling and training customization.