This product was not featured by Product Hunt yet. It will not be visible on their landing page and won't be ranked (cannot win product of the day regardless of upvotes).
Product upvotes vs the next 3
Waiting for data. Loading
Product comments vs the next 3
Waiting for data. Loading
Product upvote speed vs the next 3
Waiting for data. Loading
Product upvotes and comments
Waiting for data. Loading
Product vs the next 3
Loading
TinyDetect
AI text & image detection that runs in your browser
Paste some text or upload an image, and TinyDetect estimates whether it was made by AI. The models run in your browser, so nothing you check is sent to a server. It's free and you don't need an account. Detectors can get things wrong, so treat a high score as a reason to take a closer look.
Most AI detectors ask you to upload whatever you're checking. That's a problem when it's a student essay, a job application or a personal photo, so I wanted to see if the whole check could run on your own computer instead.
I didn't train the detectors myself. TinyDetect uses two existing open models: Fakespot's RoBERTa model for text and the Community Forensics model for images. My work was making them small enough to load in a browser. The text model went from about 500 MB to 97 MB, and the image model is about 26 MB. Both run with ONNX Runtime Web, which uses your GPU if your browser supports it.
Nothing you paste or upload leaves your browser. The page still has to download the models, the runtime and a few fonts, so the first visit takes a while. After that everything is cached, and a text check takes about 60 milliseconds on my laptop.
A few things to keep in mind:
- Detectors make mistakes, so please don't use a score to fail a student or turn down a job candidate.
- Short text, heavily edited writing, screenshots and cropped images are harder to judge.
- Phones may have trouble loading the models.
- Right now the web checker is the only part that's live. The browser extension, Gmail and Docs add-ons and SDK on the site are still in the works.
I'd really like to hear where it gets things wrong, and whether checking things locally would help in your work.
About TinyDetect on Product Hunt
“AI text & image detection that runs in your browser”
TinyDetect was submitted on Product Hunt and earned 2 upvotes and 1 comments, placing #153 on the daily leaderboard. Paste some text or upload an image, and TinyDetect estimates whether it was made by AI. The models run in your browser, so nothing you check is sent to a server. It's free and you don't need an account. Detectors can get things wrong, so treat a high score as a reason to take a closer look.
On the analytics side, TinyDetect competes within Education, Privacy, Artificial Intelligence and GitHub — topics that collectively have 613.9k followers on Product Hunt. The dashboard above tracks how TinyDetect performed against the three products that launched closest to it on the same day.
Who hunted TinyDetect?
TinyDetect was hunted by Ritabrata Maiti. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
For a complete overview of TinyDetect including community comment highlights and product details, visit the product overview.
Ritabrata Maiti
Hi everyone, I'm Ritabrata and I made TinyDetect.
Most AI detectors ask you to upload whatever you're checking. That's a problem when it's a student essay, a job application or a personal photo, so I wanted to see if the whole check could run on your own computer instead.
I didn't train the detectors myself. TinyDetect uses two existing open models: Fakespot's RoBERTa model for text and the Community Forensics model for images. My work was making them small enough to load in a browser. The text model went from about 500 MB to 97 MB, and the image model is about 26 MB. Both run with ONNX Runtime Web, which uses your GPU if your browser supports it.
Nothing you paste or upload leaves your browser. The page still has to download the models, the runtime and a few fonts, so the first visit takes a while. After that everything is cached, and a text check takes about 60 milliseconds on my laptop.
A few things to keep in mind:
- Detectors make mistakes, so please don't use a score to fail a student or turn down a job candidate.
- Short text, heavily edited writing, screenshots and cropped images are harder to judge.
- Phones may have trouble loading the models.
- Right now the web checker is the only part that's live. The browser extension, Gmail and Docs add-ons and SDK on the site are still in the works.
I'd really like to hear where it gets things wrong, and whether checking things locally would help in your work.