GTO Wizard has launched preflop ICM solving, and the headline number is the one that matters: spots that previously took around 30 minutes to solve now return an answer in about one second. For tournament players, that is not an incremental upgrade to a study tool. It changes what is practical to study at all.
The feature is available to Ultra subscribers and, critically, runs without local hardware — you can solve a nine-handed ICM spot on a phone.
What actually shipped
The core capability is a customisable preflop tree solved under ICM rather than chip EV. Users set stacks, antes, bounties and payouts, then rebuild any tournament hand in either model.
The specifications:
- Support for up to 9 players at the table
- Stacks up to 250 big blinds
- Customisable raise sizes and nodelocking
- A seamless preflop-to-postflop transition inside the same tool
- Field sizes supported up to 4,096 entrants
- Fully customisable payout structures
Format coverage is broad: standard MTTs, satellites, knockout, progressive knockout (PKO), total knockout and mystery bounties. Satellites and mystery bounties are the two formats where intuition fails hardest and where a solver most obviously earns its keep — satellite bubbles because chip value collapses to a step function, mystery bounties because the prize distribution changes correct stack-off ranges in ways that are genuinely hard to eyeball.
There is also hand history upload integration. You import stacks, antes and bounties straight from a hand history, set payouts and the number of players remaining, and solve. That removes the data-entry step that made ICM study tedious even before the solve time.
The workflow problem this solves
Per GTO Wizard's own blog, the previous process for a single preflop ICM spot meant laborious manual entry followed by roughly a half-hour wait. Change one variable — a stack size, a payout — and you waited another half hour.
That is not a tool people use to study. That is a tool people use to settle an argument about one hand, once. The practical consequence was that most tournament players studied ICM through pre-solved charts and heuristics rather than by interrogating their own spots, which meant their understanding of why a range shifted was second-hand.
High-stakes tournament player Samuel Muller put the change in workflow terms: "For me, the biggest thing about this release is efficiency. Until now, studying a specific ICM spot often meant running the preflop sim, waiting for it to solve, and then moving between different tools for the rest of the analysis. Being able to do all of that directly in GTO Wizard is now saving me a lot of time."
The second-order effect is more interesting than the time saved. When a solve costs one second instead of 30 minutes, you stop asking "what is the answer here?" and start asking "how does the answer move when I change this?" Sensitivity analysis — sliding a stack from 18bb to 22bb and watching a shoving range breathe — is where real intuition gets built, and it was previously unaffordable.
The limitation worth knowing about
Preflop trees currently allow up to three players to reach the flop. This is the same constraint GTO Wizard applies to its multiway preflop solving for cash games, and it exists for computational reasons: each additional player to the flop multiplies tree size dramatically.
In practice, spots where a fourth player over-calls or over-limps are not yet solvable. GTO Wizard has said this is on the roadmap as one of the next features.
How much this matters depends on your games. In tough mid-to-high-stakes MTTs, four-way flops are uncommon enough that the limitation is minor. In low-buy-in fields with heavy limping, four-way pots are routine — though the ICM implications of a limped four-way pot are also rarely the spots costing those players money.
Pricing, and the timing of it
GTO Wizard is running an early-bird offer with a deadline of 1:00 p.m. CEST on October 15, 2026, open to both monthly and annual subscribers:
- Monthly Ultra: $279, down from $359
- Annual Ultra: $229 per month, down from $289
The company's own framing is that locking an annual plan before the deadline saves at least $720 across 12 months, and that switching from monthly to annual before October 15 saves $1,560 over the following year.
Those are serious numbers, and they deserve a serious caveat. At $229 a month, a subscription costs $2,748 a year. For a player grinding $22 and $55 MTTs, that is a meaningful chunk of an annual bankroll and almost certainly not the highest-EV expenditure available. The players for whom Ultra pricing is obviously correct are those regularly in four- and five-figure buy-ins, where one better ICM decision at a final table pays for the year.
It is also worth noting that PokerNews' coverage of the launch is partner content containing affiliate links, with no independent testing or third-party benchmark cited. The specifications come from GTO Wizard. That does not make them wrong, but the one-second solve claim has not been externally verified.
What this means for players
If you play tournaments seriously, ICM is the highest-leverage study area available. Most players' preflop chip-EV ranges are broadly fine. Their final-table and bubble adjustments usually are not, and the money in tournaments is concentrated exactly there. If you are not yet clear on why, start with our explainer on what ICM is before spending anything on a solver.
Cheap solves change how you should study. With 30-minute solves you memorised outputs. With one-second solves you should be running the same spot at five different stack depths and watching what moves. The generalisable lesson — "my calling range collapses fastest when there are two shorter stacks behind" — is worth more than any individual chart.
Match the tool to the stakes. A $2,748 annual subscription needs to return several thousand dollars in improved decisions. Below about the $500 buy-in level, free resources plus disciplined hand review will get you most of the way. Our poker strategy hub, the poker odds fundamentals and a free poker odds calculator cover a lot of ground at no cost.
Beware the deadline pressure. An October 15 cutoff attached to a $1,560 saving is a sales mechanism, and the correct response to a sales mechanism is to decide whether you want the product first and check the price second.
The solver arms race is not slowing. GTO Wizard shipped a PLO solver update in July and this in September, and a competing ICM course from Alex Theologis launched through Run It Once earlier this month. The edge is not in having access to solutions — everyone at mid-stakes and above does now. It is in the speed and accuracy with which you convert solutions into table decisions, which remains a human skill.
For players wanting to put ICM work into practice, the tournaments hub covers where the volume is, and our cash games vs tournaments comparison is worth reading if you are not certain tournaments are where your edge lies in the first place.
Sources: PokerNews