ZenoSolver: Holdem & OmahaCustom preflop solving for Tournaments and Cash
Enter a table's players (up to 9), stacks, antes and, for a tournament, its payouts, and solve it. Then study every preflop decision, with ICM for tournaments, the same way you study a library tree.
Hi, I'm Oleg Ostroumov.
I've built private mixed-game solvers for players like Jungleman and Trueteller.
Now I'm building Zeno: a Holdem and PLO solver for everyone. It's in beta, free to try, and I'm actively developing it, as of October 2026.
Where it is today:
- Holdem flops in 2 seconds to 0.1 % of the pot exploitability
- Custom preflop solving for Tournaments and Cash
- Automatic bet sizing
- Card bunching: folded players' cards change the deck
- Preflop Node-locking
- Postflop Node-locking, including aggregate frequencies across all rivers
- River reports
- Play against the solver
- PLO
I'd love to hear what you think, and what would make Zeno useful for you: email / Telegram
Real-time recording of the live site: Holdem 6-max 100bb, button vs big blind, flop Q♠ 9♥ 4♦, 4 bet sizes.
New in Zeno
The 2026 WSOP Main Event final table with 7 left, at the start-of-day stacks. Folded to the cutoff with 12 bb: ICM raises A♥ 6♣ to 2 bb or folds it, and never moves all-in.
Preflop node-locking
Lock a preflop node of the 6-max cash trees: a frequency for the whole range, or a strategy for the hands you pick, and re-solve. Pinned hands carry a yellow dot, and one switch flips every node between Node-locked and GTO.
Big blind vs the button's 2.5 bb open: fold 70 %, and always 3-bet T8s 97s 86s 75s 64s 53s. Re-solved, the button opens 47.0 % instead of 41.5 %.
Auto sizing. For every bet and raise, Auto picks the size between 20 % and 150 % of the pot that gives up the least EV in that exact spot. It looks at the board, both players' ranges (how strong, how much air), the stack-to-pot ratio, the pot type and who raised preflop — a model trained on solves that tried every size. It adds a second size only if that gains at least 0.05 % of the pot; otherwise one size, so the tree stays small and the solve fast. All-in is always available.
Three things other solvers don't do
Card bunching
When players fold before the flop, their hands are not random, so the remaining deck contains more high cards on average. This is called bunching effect.
Zeno can solve a flop both with and without bunching effect, and you can compare solutions side-by-side.
In 6-max, not accounting for bunching effect introduces 1-1.5% pot exploitability.
9-max, button vs big blind 3-bet pot, 6♦ 5♦ 2♣. With seven players folded, the 3-bettor c-bets more on low boards: 39 % → 44 % of its range.
Node-locking, including frequencies across all rivers
You can lock an action frequency averaged over every river card, for example “IP does 3rd barrel with 85% probability across all rivers”. In the clip, Zeno re-solves in 2.8 s and hits the 85% target exactly.
Button vs big blind, K♠ 7♦ 2♣ 6♥, the button's third barrel: River report → lock icon → 85 % → Lock + re-solve.
River reports
View the strategy on all rivers at once.
Holdem 6-max 100bb, J♥ 9♥ 4♣ 2♠: every river at once.
Also: Play mode
Play against the solver. See solver frequencies and EV loss for each action. See the strategy at that spot in one click.
A mistake on the flop: an all-in with jacks that the solver never makes, −1.15 bb.