AI · alternatives · 2026
24 alternatives to WaifuXL - SOTA image upscaling neural network that runs in the browser
WaifuXL is our state of the art super resolution model trained on ~1,200,000 anime style images from the Danbooru2021 dataset. You can run it straight from your browser, entirely client-side, at https://waifuxl.com. Using the Onnx Runtime, we stream the weights of our ML models directly onto your device to be executed locally in WebAssembly. Doing so allows us to provide this service solely through a static webpage, no backend for model execution needed. This has the added benefit of enabling the privacy of your images--your images are not, and never will be, sent to us. Our…
WaifuXL is a neural network for image upscaling trained on 1.2 million anime-style images. It runs entirely in the browser using WebAssembly, with no server backend required, meaning images never leave the user's… Below are 24 products that do a similar job, ranked by how close each is in meaning and then by launch-day votes. WaifuXL - SOTA image upscaling neural network that runs in the browser launched in 2022; newer entries below may have overtaken it.
- 1
- 2
A state-of-the-art AI that draws custom anime portraits, just for you! This machine learning artist figures out your preferences and creates a perfect character illustration in 4 easy steps. If it sounds like magic, that's because it is!
2019 · waifulabs.com · its alternatives →
- 3
V2Fun▲449Generate 3D character with 8K textures and AI motion capture
Jul 2026 · v2fun.ai · its alternatives →
- 4
Image Upscaler▲235Upscale and enhance your photo, animes and cartoons in batch
2021 · imgupscaler.com · its alternatives →
- 5
Precise element control in v7 images
2025 · updates.midjourney.com · its alternatives →
- 6
PerceptPixel 2.0▲203Canva and Cloudinary on a jetplane to Milan
2024 · perceptpixel.com · its alternatives →
- 7
MAI-Image-2.5▲241Hey HN I’ve been experimenting with different image generation tools for months — but I kept running into the same problems: slow generation, repeated outputs, and limited artistic control. That’s why I built MAI-Image-1, Microsoft’s latest AI image generation model designed for creators, designers, and developers who need fast, realistic results without sacrificing flexibility. What makes MAI-Image-1 different: Ultra-realistic rendering — it accurately captures light, texture, and natural shadows. Lightning-fast generation — create detailed images in seconds. No style lock-in — experiment…
Jun 2026 · microsoft.ai · its alternatives →
- 8
- 9
Hi HN, author here. SHARP is Apple's recent single-image 3D Gaussian splatting model (https://arxiv.org/abs/2512.10685). Their reference code is PyTorch + a pretty heavy pipeline; I wanted to see if it could run in a browser with no server hop, so I exported the predictor to ONNX and ran it via onnxruntime-web with the WebGPU EP. What works: drop in an image, get a .ply you can download or preview live, all on your machine — your image never leaves the tab. The model is large (~2.4 GB sidecar) so first load is slow on a cold cache, but inference itself is a few seconds on…
May 2026 · github.com · its alternatives →
- 10RA
Hey HN! My friends and I built a new platform for generating images. The app is easy to use for people who find ComfyUI hard to use, or just simply don’t have a GPU to use it on. For those not familiar, ComfyUI is a great tool for using open-source models like Stable Diffusion. It’s primarily great because it’s a node-based tool, which means you can chain together models, upscalers, prompting nodes, etc… which let you create images in the exact aesthetic you want. There’s also a vibrant dev community behind ComfyUI, which means that there are a ton of nodes and customizability. We’re users…
2024 · its alternatives →
- 11
A private AI companion that lives on your computer,
Jun 2026 · localwaifu.com · its alternatives →
- 12
- 13
- 14
Hey HN, My wife is a designer, and one day she asked me if there was a tool to enhance images and improve their quality. I recommended a few to her, but she found that the suitable ones were too expensive, and the more affordable options didn't offer enough magnification, most only going up to 2x or 4x. So, I thought, why not create a tool that meets her needs, one that can greatly enhance image resolution and clarity while being very affordable to use? After over a month of effort, I finally completed this tool. It can now enlarge images up to 10 times, with a maximum support of 12,000…
2024 · mejorarimagen.org · its alternatives →
- 15IT
I built this app over a week as a way to learn how to use TensorFlow with Mobilenets and to get some experience with Google Play (and partly as a dare). It's written in Java, as I wasn't able to find a Kotlin API for TFLite. It was built with Bazel. I'm pretty satisfied with the actual detector's performance, although I expect I could improve the UI a little bit. It's a weird UX, as you want it to be as simple and fast as possible (loss/notloss) but you also want it to have some sort of recognizability. I would love to hear your thoughts.…
2018 · its alternatives →
- 16
Implementation of some interesting ideas of deeplearning. - DeepLearningForFun/Mxnet-Scala/Pix2Pix at master · Ldpe2G/DeepLearningForFun
2017 · github.com · its alternatives →
- 17
The Emotion Engine has 32 MB of RAM total, so the trick is streaming weights from CD-ROM one matrix at a time during the forward pass — only activations, KV cache and embeddings live in RAM. This means models bigger than the RAM can still run, they just read more from disc. Had to build a custom quantized format (PSNT), hack endianness, write a tokenizer pipeline, and most of the PS2 SDK from scratch (releasing that separately). The model itself is also custom — a 10M param Llama-style architecture I trained specifically for this. And it works. On real hardware.
Mar 2026 · github.com · its alternatives →
- 18
It's our new text-to-image model: a 9.3B single-stream diffusion transformer trained entirely from scratch. We focused heavily on controllability through structured JSON prompts, with strong text rendering, spatial awareness through bounding box guidance, and color palette control. It has the best text rendering of any open-weight model we've tested so far, and the NF4 quantized checkpoint runs on a single 24GB GPU. For more technical details and examples see our blog post: https://ideogram.ai/blog/ideogram-4.0/ We will be happy to answer any questions :)
Jun 2026 · github.com · its alternatives →
- 19
Synthetic Anime-Style Female Faces
2019 · thiswaifudoesnotexist.net · its alternatives →
- 20
- 21RT
The story of Vectorly — four pivots, three years, and a Hopin acquisition. From video compression to AI upscaling to virtual backgrounds. Founded 2018, acquired by Hopin December 2021.
2021 · vectorly.io · its alternatives →
- 22IC
Hi, my name is Rayn. I am a solo independent entrepreneur. Initially, my only reason for learning was to make tons of money, but after completing the learning process, I was very confused about what my first project on the internet should be. I started browsing the internet for ideas but got bored and wasted time on social media. I noticed that in the comment sections of anime clips, many people were asking for the anime name and episode. That's when I thought, "Let's make an anime reverse search engine," and boom, Cleithral was born. Just upload a video, image, or gif, and it will give you…
2024 · cleithral.com · its alternatives →
- 23
Online AI-powered image enhancer for upscaling anime photos
2022 · vanceai.com · its alternatives →
- 24
I’ve spent the last year developing a decomposition kernel that addresses the "Linear Bottleneck" in standard compression (Zstd, LZMA, Brotli). The Problem: Traditional algorithms operate on 1D streams. They are blind to the spatial, temporal, or hierarchical architecture of modern data (3D models, nested JSON, multi-modal blobs). When you force these structures into a 1D window, you leave significant entropy on the table. The Solution: PXG v3.0 maps binary data to a 13505x13505 grayscale grid. By treating data as a coordinate system rather than a line, our engine utilizes adaptive tiling to…
2025 · limitpixel.com · its alternatives →
Also compare
Ranked by how close each launch is in meaning, then by votes. Prices were read from each product’s own site when checked and can change. Refine with your own description →