https://github.com/philz1337x/clarity-upscaler
Clarity AI | AI Image Upscaler & Enhancer - free and open-source Magnific Alternative
https://github.com/philz1337x/clarity-upscaler
ai ai-art image-upscale image-upscaler image-upscaling image2image img2img stable-diffusion stable-diffusion-webui upscale upscaler upscalerimage
Last synced: 6 days ago
JSON representation
Clarity AI | AI Image Upscaler & Enhancer - free and open-source Magnific Alternative
- Host: GitHub
- URL: https://github.com/philz1337x/clarity-upscaler
- Owner: philz1337x
- License: agpl-3.0
- Created: 2024-03-15T15:21:40.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-04-13T12:03:18.000Z (about 1 year ago)
- Last Synced: 2024-04-14T01:07:41.157Z (about 1 year ago)
- Topics: ai, ai-art, image-upscale, image-upscaler, image-upscaling, image2image, img2img, stable-diffusion, stable-diffusion-webui, upscale, upscaler, upscalerimage
- Language: Python
- Homepage: https://ClarityAI.cc
- Size: 30.8 MB
- Stars: 2,178
- Watchers: 27
- Forks: 230
- Open Issues: 8
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE.txt
- Citation: CITATION.cff
- Codeowners: CODEOWNERS
Awesome Lists containing this project
- AiTreasureBox - philz1337x/clarity-upscaler - 04-07_4250_4](https://img.shields.io/github/stars/philz1337x/clarity-upscaler.svg)|Clarity-Upscaler: Reimagined image upscaling for everyone| (Repos)
README
Clarity AI | AI Image Upscaler & Enhancer - free and open-source Magnific Alternative
[](https://ClarityAI.co)
[](https://ClarityAI.co/api)
[](https://twitter.com/philz1337x)
[Full Video on X/Twitter](https://x.com/philz1337x/status/1768679154726359128?s=20)
# 👋 Hello
I build open source AI apps. To finance my work i also build paid versions of my code. But feel free to use the free code. I post features and new projects on https://twitter.com/philz1337x
# 🗞️ Updates
- 03/05/2025: Flux upscaling only via my app and api: (http://clarityai.co/flux-upscaler)
- 06/19/2024: Pattern upscaling
- 05/24/2024: Increased Resolution to 13kx13k (https://x.com/philz1337x/status/1793983581636690379)
- 05/16/2024: Output file format: jpg/png/webp (https://x.com/philz1337x/status/1791431093641457824)
- 05/02/2024: Sharpen image
- 05/07/2024: ComfyUI node (https://x.com/philz1337x/status/1787905308439826920)
- 04/12/2024: Multi-step upscaling (https://x.com/philz1337x/status/1785269458304442565)
- 04/07/2024: Resemblance fixed (https://x.com/levelsio/status/1776729356120797265)
- 04/05/2024: Speed Improvements (https://x.com/philz1337x/status/1776121175195975888)
- 04/01/2024: Support custom safetensors checkpoints (https://x.com/philz1337x/status/1774772572632338435)
- 03/28/2024: Anime upscaling (https://x.com/philz1337x/status/1773342568543346738) (https://clarityai.co/anime-image-upscaling)
- 03/26/2024: LoRa Support (https://x.com/philz1337x/status/1772575319871959180)
- 03/21/2024: Pre downscaling (https://x.com/philz1337x/status/1770680096031961351)
- 03/18/2024: Fractality (https://x.com/philz1337x/status/1769756654533485050)
- 03/15/2024: Code release (https://x.com/philz1337x/status/1768679154726359128)
# 🔎 Upscaling with Flux
Flux Upscaling is now available at [ClarityAI.co/flux-upscaler](http://clarityai.co/flux-upscaler) and is not open-source
- It supports Flux LoRas with a style or a face
- It's very good at faces, text, art, and generating error-free images
# 🚀 Options to use Clarity-Upscaler
Note that this repository is an implementation for cog. If you are not familiar with cog, I recommend the easier solutions. The free options are ComfyUI and A1111, while the paid but easy-to-use options are my app ClarityAI.co and the ComfyUI API Node.
## 🧑💻 App
The simplest option to use Clarity is with the app at [ClarityAI.co](https://ClarityAI.co)
## ⚙️ API
To integrate Clarity Upscaler with an API into your application use: [ClarityAI.co/API](https://ClarityAI.co/api)
## 🐰 ComfyUI
### 1. API node
1. Open ComfyUI Manager, search for Clarity AI, and install the node.
2. Create an API key at: [ClarityAI.co/ComfyUI](https://ClarityAI.co/comfyui)
3. Add the API key to the node as a) envirement variable `CAI_API_KEY` OR b) to a `cai_platform_key.txt` text file OR c) in `api_key_override` field of the node.
Full instructions: https://github.com/philz1337x/ComfyUI-ClarityAI
### 2. Free workflow
1. Download the repo https://github.com/philz1337x/ComfyUI-ClarityAI and use the file free-wokflow.json
## 🕵️♂️ Cog
If you are not familiar with cog read: cog docs
- run `download_weights.py`
- predict with cog:
```su
cog predict -i image="link-to-image"
```
## 🤹♂️ A1111 webUI
For a detailed explanation, use the tutorial in this post: https://x.com/philz1337x/status/1830504764389380466
https://github.com/AUTOMATIC1111/stable-diffusion-webui
- Use these params:
```Prompt:
masterpiece, best quality, highres, Negative prompt: (worst quality, low quality, normal quality:2) JuggernautNegative-neg Steps: 18, Sampler: DPM++ 3M SDE Karras, CFG scale: 6.0, Seed: 1337, Size: 1024x1024, Model hash: 338b85bc4f, Model: juggernaut_reborn, Denoising strength: 0.35, Tiled Diffusion upscaler: 4x-UltraSharp, Tiled Diffusion scale factor: 2, Tiled Diffusion: {"Method": "MultiDiffusion", "Tile tile width": 112, "Tile tile height": 144, "Tile Overlap": 4, "Tile batch size": 8, "Upscaler": "4x-UltraSharp", "Upscale factor": 2, "Keep input size": true}, ControlNet 0: "Module: tile_resample, Model: control_v11f1e_sd15_tile, Weight: 0.6, Resize Mode: 1, Low Vram: False, Processor Res: 512, Threshold A: 1, Threshold B: 1, Guidance Start: 0.0, Guidance End: 1.0, Pixel Perfect: True, Control Mode: 1, Hr Option: HiResFixOption.BOTH, Save Detected Map: False", Lora hashes: "more_details: 3b8aa1d351ef, SDXLrender_v2.0: 3925cf4759af"
```