Finish a hand and know within 30 seconds where you lost money, how much, and what to do instead. An AI coach explains why, and your mistakes come back until they stop being mistakes.
Paste a hand history and every decision gets an EV loss. See which hands leaked chips and on which street.
Not just the answer — the reason. Blockers, pot odds, range advantage, explained in plain language.
Play real hands against the solution. Mistakes go into a review queue and come back until you stop making them.
Paste GG or PokerStars histories directly; log a live hand in about 20 seconds. Scan a whole session at once and get a leak report by hand class — "your top pairs lose 3.2% on average, and the problem is the turn."
This is the feature other tools still label Coming Soon.
Play complete hands against the solution: the opponent acts from its real range, cards run out to showdown, and every decision is scored as you go.
Costly decisions land in your review queue and keep reappearing until you stop repeating them. That loop is what actually moves your win rate.
Any board, any stack depth, ante and straddle structures, custom bet sizing, nodelocked exploit research, equity distribution curves — all of it.
Others put these behind $79–$139 tiers. Ours are in the $12 one.
The real risk with a solver is being fast and wrong. Here is the engine’s correctness, in the open.
580 decision nodes aligned one by one, then its strategy measured with our own independent best-response implementation — 0.31% pot exploitability. Two engines checking each other.
The clairvoyance game has a known analytic equilibrium (IP bets 0.75× pot). Started from zero, the engine converges to exactly that.
From hand enumeration and suit isomorphism to CFR convergence, every change has to pass all 41 tests before it merges.
Web, iOS and Android run the same Rust core rather than three separate implementations. Same hand, same answer, everywhere.
Anything else — write to hi@vipergto.com
A GTO (game-theory optimal) solver simulates a spot over and over until neither side can find a better line — that "nobody can gain by deviating" strategy is the equilibrium. It does not tell you this hand will win. It tells you the line no opponent can exploit over the long run, and how much EV your actual line gave up against it.
They are more expensive with bigger prebuilt libraries; we are cheaper and solve in real time. Others ration postflop solving on the free tier and lock research features like nodelock behind their top tier. We make solving and training free and unlimited, and the $12 tier already includes nodelock, batch analysis and leak reports. What you pay for is saved reviews, leak profiling and the AI coach — not how many solves you get today.
No, because it is designed from the start not to do real-time assistance (RTA). Viper GTO only works after a hand is over: you paste a history to review it, or you play against the solver in training mode. It never reads a live table, never overlays anything, and never talks to a poker client. That is a product red line, not a feature we are waiting to add.
GG and PokerStars text histories can be pasted in whole; streets, bet sizes and showdown results are parsed automatically. Live hands or other sites can be entered by hand in about 20 seconds each. A whole session can be scanned in one pass, straight into a leak report.
Postflop solving, training hands, and 10 preflop charts (RFI and BB defence) are all free and unlimited. The barrier to learning GTO should be zero. The free tier limits saved reviews and AI coach usage, because those are the parts that cost server time.
Yes. Web, iOS and Android all run the same engine. Solving happens locally on your device, so it works with no connection — reviewing last night’s hand on your commute is fine.