Jingwei vs Glaze
Visible credit on the posted file, training interference on the scrape
Jingwei and Glaze answer different fears. Glaze aims at bulk scraping and style learning after you post. Jingwei aims at someone who already has your file and wants a clean edit or a wiped mark. Artists often need both. One does not replace the other.
This page stays on the line Jingwei actually draws. For Glaze's own settings, read Glaze's docs. For Nightshade next to Glaze and a visible watermark, see Glaze vs Nightshade vs a Visible Watermark.
What Glaze is for
Glaze adds small changes meant to make style extraction from your posts less useful. People often barely notice them on the surface. Many pipelines that try to learn your look from glazed files get a noisier result.
It runs as a local app. Artists commonly need a capable GPU, more time per image than a browser watermark, and a more visible perturbation than Jingwei's light Quick Credit default. Jingwei's FAQ groups Glaze with Nightshade on the training-scrape side of the line.
Glaze does not prove you made the piece. It does not stop someone who already holds a clean original from editing that file. Protections of this kind can weaken as removal methods and new models appear. Treat it as a strong first layer against style mimicry and scrape usefulness, not a forever seal.
What Jingwei is for
Jingwei protects the export you publish. Visible layers such as displacement, emboss, halftone, and face emboss lock change pixels so a cheap inpaint tends to leave dirt. A JW claim stores your name, whether the work is Original or AI-assisted, and usage limits you state. Verify can read those Jingwei signals later.
It runs in the browser. Everyday Protect needs no account. Processing is classical image processing with no generative model in the loop. One image takes about 30 seconds. Uploads are deleted after processing.
A hit on Verify means a Jingwei signal is still readable. It is evidence and intent, not independent legal proof of ownership. Invisible layers can weaken after heavy compression, repeated saves, or a full repaint.
Side-by-side in plain terms
Main job. Glaze: make training scrapes and style learning harder. Jingwei: raise wipe and anonymous reuse cost on the file you posted, and attach a claim others can check.
Where it runs. Glaze: local client, often GPU-heavy. Jingwei: browser Protect and Verify.
Account. Glaze follows its own install flow. Jingwei everyday Protect does not need signup.
Speed. Glaze is usually slower per image. Jingwei targets about 30 seconds for a typical protect pass.
Proof of ownership. Neither tool is a court filing. Jingwei's JW claim is a readable attribution signal. Glaze is not built as a public verify page for your name.
Together. Yes. Run Glaze (and Nightshade if you use it) as their own export. Then protect the public file you are willing to show in Jingwei. Keep a copy from each step until you have checked the result.
When to pick which
Pick Glaze when your main fear is scrapers learning your style from posts. Pick Jingwei when your main fear is edits, fusion, reposts without credit, or a wiped mark on a file someone already downloaded. Pick both when you have both fears.
Inside Jingwei, Quick Credit is the daily default for a faint credit and JW claim. Manual is for stacking stronger visible layers yourself. Layer samples live on the watermark matrix.
Is Jingwei a Glaze alternative?
Only for the posted-file job. If you need training interference, keep Glaze (or Nightshade). If you need visible deterrence and a checkable claim on the upload, use Jingwei. Calling either a full substitute for the other sets the wrong expectation.
Do I run Glaze or Jingwei first?
Keep the unmarked master. If you glaze, do that on the export you accept as public-looking. Then run Jingwei on the file you will actually upload, or follow whatever order you have tested yourself. This site does not claim that Protect preserves a Glaze perturbation, or the reverse. Check your own exports.
Protect the file someone can already download
If the next post is what you care about, glaze when training is the fear, then protect the public export when edits and credit are the fear.
Open Protect on Jingwei