The TableLab GTO Library

TableLab's Study tab holds 26,325 fully solved postflop spots — five common preflop battles, every one of the 1,755 strategically distinct flops, three stack depths each, solved from the flop through the river. It is hosted, streamed on demand, and free inside the free TableLab app: in your browser at tablelab.app or on Android via Google Play.

What's inside

The library covers five heads-up scenarios. Each one is solved on all 1,755 strategically distinct flops and at three stack-depth buckets, which is where the 26,325 comes from: 5 scenarios × 1,755 flops × 3 depths.

The five solved scenarios. Every scenario covers all 1,755 strategically distinct flops.
Scenario Stack-depth buckets
BTN opens vs BB call — single-raised pot Shallow · Medium · Deep
CO-class (middle-position) open vs BB Shallow · Medium · Deep
UTG-class (early-position) open vs BB Shallow · Medium · Deep
SB opens vs BB call — blind vs blind Shallow · Medium · Deep
BB 3-bets vs BTN open — 3-bet pot Committed · Shallow · Medium

Every solved spot carries the full strategic detail, not just a summary:

Across the library that adds up to hundreds of millions of solved strategy datapoints.

Every flop — literally

There are 22,100 ways to deal three cards from a 52-card deck. But suits have no inherent value in hold'em: A♠7♥2♦ and A♥7♦2♣ are the same strategic problem with the suit labels swapped. Grouping flops that are identical up to a relabeling of suits — suit isomorphism — collapses those 22,100 raw flops to 1,755 strategically distinct ones. The library solves each of the 1,755 once, represented by one canonical suit assignment, so every possible flop you can be dealt maps to a solved board. That is what "every flop" means here: complete coverage, no sampling, no "similar texture" approximation at the flop level.

How it was built

Every spot was solved with TexasSolver, an open-source CFR (counterfactual regret minimization) solver that TableLab uses under a commercial license, running on large cloud compute over multiple solve campaigns. Each solution is converged to an exploitability of at most 0.5% of the pot, and the full game tree is solved from the flop through the river — turn and river strategy comes from the same converged tree, not from re-solving or extrapolation.

What ≤0.5% exploitability means. Exploitability measures how far a strategy sits from perfect equilibrium: it is the maximum a theoretically perfect opponent could win against it, per hand, beyond what they'd win at true equilibrium. At 0.5% of the pot, the computed strategy gives up at most half a percent of the pot to a perfect counter-strategy — a standard precision target for practical study, and far below what any human opponent can extract.

Browsing it: the explorer

The Study tab is a visual explorer over the solved trees, built in the style players already know from modern GTO trainers:

Free, inside a free tracker

Precomputed GTO solutions have generally been sold as subscriptions: PeakGTO lists its paid plan at $79 per month, Lucid's Pro tier runs $49 per month, and GTO Wizard's paid tiers span $49 to $279 per month billed monthly ($39 to $229 on annual billing), per PokerNews's 2026 pricing overview — all as published in August 2026. Those are capable products built for dedicated solver study. TableLab takes a different position: the library ships free inside a free bankroll tracker, because the point is not to sell solver access — it is to put a solved baseline next to the sessions and hands you already track.

Connected to your own hands

The library isn't a standalone viewer bolted onto the app. TableLab records your sessions and full street-by-street hand histories, and its AI hand coaching is grounded in deterministic facts — computed equities, pot odds, and solver-derived GTO frequencies from this same library — injected into every analysis so the coaching can't invent numbers. Study a spot in the explorer, then see the same solved frequencies referenced when the coach reviews a hand you actually played.

Honest limitations

Open the library. The Study tab is live in the free TableLab app — open the web app or get it on Google Play. No signup fee, no locked scenarios.

FAQ

What is GTO poker strategy?

GTO (game-theory optimal) strategy is the equilibrium solution to a poker spot: a mix of bets, checks, calls, raises, and folds for every hand such that no opponent strategy can exploit it. Solvers compute these equilibria numerically. Studying GTO output tells you the baseline — how often each hand in your range wants to take each action — which you can then deviate from to exploit real opponents.

Is the TableLab GTO library really free?

Yes. The library is part of the free TableLab app — the web app at tablelab.app and the Android app on Google Play. There is no paywall on the Study tab, no locked scenarios, and no credit card. Comparable precomputed-solution products are paid subscriptions.

How accurate are the solutions?

Every spot is solved with TexasSolver, a CFR solver used under a commercial license, to an exploitability of at most 0.5% of the pot. That means a theoretically perfect opponent could gain no more than half a percent of the pot per hand against the computed strategy — a standard precision target for practical GTO study. The in-app footer labels the library Beta, flop to river.

Why do the suits in the library sometimes differ from my actual hand?

Boards are stored in canonical suit-isomorphic form. Suits have no inherent value in hold'em, so A♠7♥2♦ and A♥7♦2♣ play identically — the 22,100 raw flops collapse to 1,755 strategically distinct ones. The library solves each distinct flop once and represents it with one canonical suit assignment; map your hand's suits onto the canonical board and the strategy is exactly the same.

Does the GTO library work on mobile?

Yes. The library lives in the TableLab app's Study tab, which runs on Android via Google Play and in any modern browser at tablelab.app — the same solved spots, streamed on demand, on both platforms.

Can it solve my exact custom spot?

No — it is a precomputed library, not an on-demand solver. Solutions exist for five specific preflop lines with standard open sizes, across three stack-depth buckets per scenario. If your spot came from a different preflop line or a non-standard size, the nearest library spot is a reference point, not an exact answer; a custom tree needs a full solver.

Solver output is a strategic baseline, not a guarantee against real opponents — exploiting the player in front of you usually beats equilibrium. Also free: the variance calculator and the ICM deal calculator.