All AI Tools
About these tools
Prompt writing is mostly answering the questions a model would otherwise have to guess at: who the output is for, what tone it should take, how long it should be, and what format you want it back in. Leave those unanswered and the model picks for you, which is why a vague prompt gives a vague answer. This tool turns those questions into fields and assembles the result for you — entirely from templates running in your browser, so it is instant, has no usage limit and costs nothing to run.
It deliberately does not call any AI service. That keeps it free and private, but it also sets the boundary of what it can do: it builds the instruction you send to a model, it does not generate the answer. You take the result to whichever assistant you already use.
Quezlo also runs a genuine AI model elsewhere on the site — the background remover loads an open-source model into your browser and cuts out a photo's subject on your own device, without the image ever being uploaded. It sits under Image Tools because that is where people look for it, but it works on the same principle: the model runs where you are, not on a server.
Nothing here needs an account, and nothing you put into any of them is uploaded or stored. Every tool is a single page that does one job, so you can bookmark the one you use and skip this page entirely next time.
What "AI" means on this site
The term covers a great deal of very different things, so it is worth being specific. The background removal tools here run a real machine-learning model — an open segmentation model that separates a subject from its background — and it runs inside your browser on your own device rather than on a server. The first use downloads the model files, which is why there is a wait, and afterwards they are cached and the tool works offline.
The prompt builder is not a model. It is a structured way of assembling a well-formed prompt from the elements that reliably matter — subject, style, composition, lighting, detail — for use in whichever image or text generator you already have. It generates a prompt, not a picture.
Why in-browser matters here
Background removal is a good example of the difference. Hosted services upload your image to their servers, which for a product photograph, a client's artwork or a picture of a person is a meaningful decision rather than a technicality. Running the model locally means the image never leaves your device, and it also means there is no per-image limit, no account and no queue, because there is no server cost to recover.
The trade is that your device does the work. A recent browser is required, results appear more slowly on older hardware, and in-app browsers inside social media apps often cannot run these models at all. If a tool will not start, opening the page in a normal browser resolves it most of the time.
What to expect from the results
Automatic background removal handles clear subjects against distinct backgrounds very well and struggles with the cases that are genuinely hard: fine hair, motion blur, transparent objects such as glass, and subjects whose colour closely matches what is behind them. Those limits are properties of the task rather than of this particular implementation, and hosted commercial services hit the same walls with the same images. Where a result is imperfect, a manual touch-up in an editor afterwards is usually faster than trying repeatedly for a clean automatic cut.