Resy Sniper Apps and Reservation Bots, Explained

Yes, Resy sniper apps are real, some of them genuinely work, and the reason they exist is simple arithmetic: at a hard restaurant, a released table is claimed in seconds, and a human checking a phone a few times a day is competing against software that checks every few seconds and never sleeps. The category is messier than it looks from a search results page, though. The phrase "reservation bot" covers at least four different kinds of tool, and they differ enormously in how well they work, what they cost, and how much risk they put on your account. Some are notification services that tell you a table appeared and leave the sprint to you. Some book automatically on your own account. Some are scripts you run yourself. And some are resale marketplaces that are not booking tools at all, they are scalping with a nicer interface.

The single most useful question to ask about any of them is this: whose account does the reservation end up on, and whose name is on the table? A tool that watches availability and books on your own account, in your own name, is doing an attention job a person cannot do, and nothing about the resulting reservation is unusual from the restaurant's side. A tool that books under other identities and transfers or sells the table to you is a different business entirely, one that restaurants and, in New York, lawmakers have moved against. Everything below is current as of September 2026, and the mechanics of why tables vanish this fast in the first place are covered in why Resy tables vanish instantly.

The four kinds of tool people call a "Resy bot"

Notification services. The oldest and most common kind, including Resy's own built-in Notify feature. They watch a sold out date and alert you when a table opens. The honest limit: an alert is a starting gun, not a table. At a genuinely hard restaurant the seat is often gone before you have unlocked your phone, because everyone else on the list got the same ping, and other software got there without needing a ping at all. That failure mode is common enough to have its own page, why Resy Notify never gets you the table.

Auto-booking watchers. The step past notification: you connect your own reservation account, describe the table you want (restaurant, dates, party size, time range), and the service books it the instant one appears, on your account, in your name. SeatSwiper is in this category. The advantage over an alert is the whole point of the category: the seconds between "a table exists" and "a table is held" are the entire game, and software spends them booking while a human spends them fumbling for a phone.

DIY scripts. GitHub has a long tail of open source Resy and OpenTable booking scripts. Free, and some work, but you are signing up to maintain them: platforms change their sites and their defenses regularly, a script that ran in March quietly dies in April, and an aggressively configured one that hammers the site every second is exactly the traffic pattern platforms build detection for. If you are not comfortable reading and fixing the code yourself, this lane costs more time than it saves.

Reservation resale marketplaces. Sites and brokers where you buy a reservation someone else already holds, sometimes for hundreds of dollars at the hardest rooms. These are not sniper apps, they are the demand side of sniping done at industrial scale: inventory is often farmed by bots under throwaway or borrowed identities and then flipped. Restaurants dislike them intensely, because a resold no-show is an empty table the house cannot fill, and because the name on the book is not the person walking in. New York moved against the practice with a reservation anti-piracy law targeting third party resale without the restaurant's agreement. Buying a resold table also carries a practical risk the listing never mentions: a reservation under a stranger's name can be cancelled the moment the restaurant notices, and some hard rooms now check.

What actually works

Two moments produce tables at a sold out restaurant, and good tools play both. The first is the scheduled release, when a new date enters the booking window at the hour the house sets, most famously the 10am convention covered in the 10am drop explainer. Software helps here, but the release is a synchronized race with many entrants and few winners, and a bot can lose it like anyone else. The second moment is the cancellation, and this is where automation stops being an edge and becomes the whole strategy. Plans collapse at random hours: a Friday 8pm comes back at 11:40 on a Tuesday night, or at 2pm the afternoon of. No one is looking then. A watcher that checks continuously wins those moments almost by default, because its only competition is other software and luck. The rhythm of when tables come back, and why deposit policies shape it, is in restaurant cancellation policies.

The corollary is worth stating plainly: a tool's win rate depends more on which moment it plays than on how fast it is. Racing releases is glamorous and low percentage. Watching for cancellations is boring and effective. Ask any service you are evaluating which of the two it actually does.

What gets accounts banned

The platforms' terms of service restrict automated access, and as of September 2026 that has been true for years, so everything in this category operates in a gray zone. In practice, enforcement lands on behavior that looks abusive rather than on the existence of automation, and the patterns that draw it are consistent. Handing your password to a service that logs in as you from unfamiliar infrastructure can trip security checks and lock the account. Scripts that poll too aggressively get their traffic blocked and can take the account down with them. Holding multiple reservations across accounts, booking under fake names, and reselling tables are the behaviors the platforms and restaurants actively hunt, because they damage the restaurant directly. And restaurants themselves are a second enforcement layer: a hard room that sees the same name booking and cancelling constantly, or a table booked under a name that does not match the diner, can simply blacklist the guest. The pattern in all of it is that risk tracks dishonesty. One diner, one table, their own name, on their own account, is hard to distinguish from an extremely attentive human, because functionally that is what it is. Farms of accounts hoarding inventory for resale are what the defenses exist to catch.

Questions to ask before trusting any of these tools

• Whose account and whose name? Your own account and your real name is the clean version of this category. Anything transferred or resold is the version restaurants cancel and New York legislated against.
• When do you pay? A success fee only when a table is actually landed aligns the tool with you. A subscription collects whether you ever eat or not.
• Which platforms does it cover? Resy, OpenTable and SevenRooms cover most of the hard-table landscape, but not all of it. Tock in particular runs prepaid ticketed bookings and is a separate world; a Resy sniper does nothing there.
• Does it watch cancellations or just race releases? The cancellation watch is where the odds are. A tool that only fires at 10am is playing the lottery with better reflexes.
• Is it a guarantee? No honest tool claims one. When one seat opens and multiple watchers want it, someone loses. A service promising certainty at any specific hard restaurant is overselling.

Where SeatSwiper sits in this landscape

SeatSwiper is an auto-booking watcher, the second category above, built deliberately on the one-diner-one-table model: it watches the specific table you describe on Resy, SevenRooms or OpenTable, around the clock, and books it the moment one opens, whether that is a scheduled release or a 1am cancellation. It books on your own account, in your own name, so there is nothing to transfer and nothing for a restaurant to object to. It does not hoard, resell, or hold multiple options, and it does not cover Tock. The first successful booking is free, after that it is a flat $5 charged only when it actually lands you the seat. It can still lose a synchronized race, like anything else in this category, but it never stops watching, which is the part no human can do. The full mechanics are in how SeatSwiper works.