A real person reads every message. Tell us what happened and we will answer.
It is shrunk on the phone, sent to our server over an encrypted connection, and passed straight to a vision model so it can be identified. We do not store it: no database, no file, nothing to look through afterwards. The privacy page lists everything that is sent and everything we record.
It will sometimes be. The model gives its best guess with a number saying how sure it is, and a low number is the app telling you to doubt it. A close photo of a single leaf, flat on and in daylight, is usually the difference. Never use the answer to decide whether something is safe to eat or touch.
There is a free allowance each day and it resets. The count is kept against a random identifier made up on your phone, not against you.
No. There is no sign in and nothing to create. The app works the moment it opens.
That is our side, not yours. We ask a second provider automatically when the first one fails, so if you see this message both were unreachable. Try again in a moment, and write to us if it keeps happening.