Revelei
Most AI photo tools upload the picture and hand back the result. Revelei runs the models on your own machine, on the GPU through DirectML or on the CPU: Real-ESRGAN to upscale 2× and 4×, GFPGAN to restore faces, RMBG with ViTMatte to cut out the background, and SCUNet to remove noise. Every model is checked against its SHA-256, and the photo never leaves the PC.
Public numbers
Install
What comes next
From the product's own docs/ROADMAP.md at b29d4e2.
Delivery: README, benchmarks and installers
- DesignedEvery published release is still PhotoBoost, and the README sends people to a Revelei installer that does not exist
- DesignedThe README and the site describe four photo operations and never say Revelei can create an image from a prompt
- DesignedA visitor who likes Revelei never learns that Modelado turns a photo into a 3D model the same offline way
- DesignedEvery first download meets a SmartScreen warning, and the Store, where photo apps are looked for, has no Revelei
- PartialNo MSIX package exists, so there is nothing to upload to the Store and nothing to install locally to test that path
- DesignedA Store release means typing the listing, uploading screenshots and the package by hand in Partner Center
- DesignedA Store install would download the CUDA engine's native DLLs after certification, code the Store bars apps from fetching
- DesignedCreate makes whatever image a prompt asks for, with no filter and no way to report a result, which Store review checks
Model manager: catalog, updates and hardware fit
- In progressThe hardware verdict stays an estimate after a model has run here; this PC's own speed is never shown
- DesignedA model card offers only the recommended quantisation; a larger or smaller one cannot be chosen
Generate an image from a prompt
- IdeaA generation that runs out of video or system memory fails instead of retrying in a way that fits
- DesignedA generated image cannot be upscaled, edited or exported, and its prompt and seed are lost once saved
- IdeaThe installer carries no generation engine, so a fresh install could not generate even after a model is downloaded
Recently shipped
All 78 entries →- PB80 · 22ba820There is no screen to write a prompt, pick a model and size, and watch an image being generated
Create keeps the session's images with their settings and offers a first model (design recorded in
src/Revelei.App/MainWindow.Create.cs). - PB79 · 7aebd62The catalog lists no text-to-image model, so there is nothing to download for generation
Z-Image Turbo, FLUX.2 klein and Qwen-Image enter the catalog, each file pinned upstream with its hash, role and memory (design superseded: Qwen-Image 2512 became the Qwen-Image the spike ran).
- PB78 · daf787dNothing in the app can turn a prompt into pixels: no layer loads a diffusion model, reports its steps or cancels it
An engine host process generates with progress, previews and cancel; a crash fails one image only (design recorded in
src/Revelei.Generation/GenerationClient.cs). - PB77 · 764877bNo engine in Revelei can run a text-to-image diffusion model, let alone one bigger than 8 GB of video memory
stable-diffusion.cpp generates 1024 px images from C# on the 8 GB desk, a 12 GB model by RAM offload; CUDA on NVIDIA, Vulkan elsewhere (design recorded in
spike/diffusion/README.md). - PB75 · 348a8d6Models are managed from a flat text list inside About, with one Download everything button
A Models panel lists each catalog model as a card with its verdict, version and queued actions; About keeps a summary (design superseded: Check now went to PB84, the picker to PB85).
- PB74 · b2d7b9fDownloaded models can only live in %LOCALAPPDATA%, so a 15 GB model fills the system drive
A chosen models folder is saved in settings and read at launch, models move there by copy, check and delete, and a missing one is said (design recorded in
src/Revelei.Inference/ModelsFolderMove.cs). - PB73 · c32cca9A multi-gigabyte download starts without checking free disk, cannot be queued, and is read twice to hash it
Files models share are kept once by SHA-256 in .shared and removed with the last model that uses them.
- PB72 · 53e5a67Nothing says whether a model will run well, slowly or not at all on this PC before its gigabytes are downloaded
Each quantisation gets a verdict (runs well, slower, CPU only, does not fit) with its reason, and the best one is recommended (design recorded in
src/Revelei.Inference/HardwareFit.cs).