Photograph a Label. Skip the Spelling Guesswork.

Yes. Send the photo on its own and @vustbot reads the text printed on the label; ask what you want to know in the next message and it answers from what the bottle actually says — grape, region, likely style, serving notes. Free tier, no card. Two things are worth knowing. It works from the label's text rather than from the look of the bottle, so a crisp printed label — including one in a language you don't read — is the easy case, while a stylised or worn one may come back unreadable, in which case it asks for a clearer shot or the name typed out. And a partly-readable label is the case to watch: it can answer confidently about the wrong bottle, so check the name it reads back.

Because they do different jobs. @vustCalBot is tuned for one narrow task — estimating calories and macros from a photo of a meal — and it is food-only by design; it is not a wine tool and does not pretend to be. @vustbot is the general chat: it reads the text off whatever you photograph and then reasons about it, which is what a wine label needs, because the useful part is understanding an appellation and a grape rather than measuring a plate.

Whatever you ask about the bottle once the label has been read: grape variety, region and country, a likely style description (light or full-bodied, dry or sweet, tannin level), practical notes like serving temperature. It is reference knowledge answered in conversation, not a database lookup against a catalogue of bottles. What you will not get is a crowd rating or a review count, because there is no rating database behind it and inventing one would be worse than useless; the Vivino comparison page covers why that distinction matters. It also cannot know the price on the shelf in front of you or whether the shop has another vintage.

The limit is legibility rather than fame: if a small producer prints its appellation and grape clearly, that reads much like a famous château, so an obscure bottle is not automatically the hard case. The genuine hard cases are minimal-text designs, ornamental or handwriting-style lettering, and worn or curved labels photographed at an angle. When the text doesn't come through at all it says so rather than guessing at a wine; the case to watch is a partial read, which can produce a confident answer about the wrong bottle. A well-lit, straight-on shot of the front label is usually the difference, and checking the name it reads back costs nothing.

Only in the input step, and that step is the whole point of this page. If you can read and type the name, region or grape, type it — the answer is the same. The photo exists for the times you can't: an appellation in a script you don't read, a name you'd mangle in a search box, or a label you simply don't want to transcribe while a waiter waits. Same conversation, same free tier, one less thing to spell.

No, and it says so plainly rather than blurring it. Vivino's strength is a crowd-rating database built from millions of real tasters; what you get here is grape, region and style knowledge, not a crowd score, and @vustbot has no way to produce one. The Vivino-alternative comparison lays out the full trade-off — which tool is right depends on whether you want to know what the bottle IS or what other drinkers thought of it.