How Cat Says Hi works

A single photo can tell you a little about a cat, read carefully. This page explains how a reading is produced, why it stays tentative, and how the site's guides are put together.

How a reading is produced

Choosing a photo only creates a local preview — nothing is sent until you select Analyze. When you do, your browser resizes the photo to a 1,600-pixel maximum edge and sends it as a JPEG, together with your optional note, to this site's Worker. The Worker passes it to Qomza's AI service, where a model reads the image and returns a structured answer: visible traits first, then possible breed cues and everyday care ideas. The whole exchange is time-limited, and the Privacy page lists the request metadata that travels with a lookup.

Readings come from an AI model

Every reading is generated by an AI model — not measured, and not checked against a breed registry. A model can be wrong, can misread lighting or angle, and can miss what a clearer photo would show. Treat each result as a starting point for looking at your own cat, not a verdict. The fictional example on the home page is a written sample — it was not produced from anyone's submitted photo.

How uncertainty is shown

Each result opens with a status line. Most read “Visible details in this photo”; when an image leaves real doubt, the label changes to “Some details are hard to see in this photo” or “Another view may help”, and a photo that may not contain a cat is marked “The photo may show something else” with cat-specific sections hidden. Breed suggestions always appear under “Possible breed cues”, described as visual possibilities only — when nothing is strong enough, the result says so rather than guessing.

What the tool doesn't do

Cat Says Hi does not confirm breed or ancestry, assess symptoms, diagnose conditions, recommend treatment, or judge a cat's mood or health from a still image. It is not veterinary triage. Health concerns belong with a veterinarian — the vet-visit notes guide suggests what to record beforehand.

How the guides are made

The guides are written from named, authoritative sources — organisations such as Cats Protection, Cornell Feline Health Center, International Cat Care, and Cat Friendly Homes — not from the tool's outputs. Each guide keeps its sources in a “Sources and review” block beside the claims they support and shows the date those sources were last checked. Where guides include illustrations or photos, captions label them — including AI-generated illustrative images.

Corrections

We read every problem report. When a guide error is confirmed, the page is corrected and, on guides that show one, the “Sources last checked” date is updated. To report a wrong reading or a guide mistake, use the report-a-problem section on the support page.