What makes our AI tools different
Most online AI services upload your file to a cloud server, process it there and often limit free use with credits or low-resolution downloads. We take a different approach. Our Background Remover uses Google’s open-source MediaPipe image segmentation model, which your browser downloads once (about 16 MB) and caches. From then on the AI runs on your own processor.
- Private by design: your photos are never uploaded.
- No credits or watermarks: download full-resolution results as often as you like.
- Works offline after the first load while the model stays in your browser cache.
Honest limitations
On-device models are smaller than the largest cloud models. They are excellent at well-defined jobs, such as separating a person from the background, but they are not designed for every subject. Where AI is not the right fit, our tools offer a reliable non-AI alternative; the Background Remover, for example, includes a colour-based mode for products and logos on plain backgrounds.
Tips for the best AI results
- Use well-lit photos with clear contrast between the subject and the background.
- Run AI tools in an up-to-date version of Chrome, Edge, Firefox or Safari.
- Give the first run a few seconds while the model downloads; later runs are much faster.
Frequently asked questions
Do your AI tools send my data to an AI company?
No. The model runs inside your browser. The only download is the open-source model file itself; your images are not sent anywhere.
Why does the first use take longer?
Your browser needs to download and initialise the model. It is cached afterwards, so the next runs usually take a second or two.
Will you add more AI tools?
Yes. We add on-device AI tools when they can run fast and privately in a browser. Tell us which AI tool you would like to see next.