Work in progress. Public API access is coming. Play Blitzfold in the meantime
THE NEXT HAND IS YOURS TO BUILD

Poker decisions.
Your next move.

Give your poker application a decision engine. Turn game context and hole cards into hand-aware action probabilities and conditional bet sizing—with one API call.

Play Blitzfold Explore the API

Public access is in development. Experience the gameplay in Blitzfold while we build.

DECISION PREVIEW01 / FLOP
YOUR HAND
A♠
A♥
2 ♣7 ♦J ♠
A hand. A game state. A distribution.
ACTION PROBABILITIESILLUSTRATIVE
Fold
2%
Call
26%
Raise
72%
Your engine samples the action. You stay in control.
ONE DECISION ENDPOINTHAND-AWARE PROBABILITIESCONDITIONAL BET SIZINGBUILT FOR DEVELOPERS
FROM GAME STATE TO GAMEPLAY

One route.
A whole range of decisions.

For developers building Hold’em bots, simulations, and poker experiences. Bring your game engine; let the API supply the action distribution.

01 / SEND

Describe the decision.

Send the encoded public game context, preceding action history, legal actions, and the acting player’s hole cards.

02 / RECEIVE

Get a distribution.

Receive fold, call, and raise probabilities when facing a bet—or check and bet probabilities otherwise. Sizing weights are conditional on aggression.

03 / PLAY

Make it your own.

Sample an action and, for a bet or raise, a size. Your engine handles legal amounts, stack limits, and execution.

LESS PLUMBING. MORE POKER.

A clear contract.
A smarter starting point.

The equity-actions endpoint combines a learned public-context policy with equity-based allocation for the acting hand.

POST/api/equity_actions
  • Exact hole cards and optional actor-range weights
  • Action probabilities and conditional sizing distributions
  • Optional tight/loose and passive/aggressive styles
JAVASCRIPT & TYPESCRIPT SDK

Our server-side SDK is being updated for the new API. The existing SDK requires Node.js 20+; public installation instructions and equity-actions support are coming with the release.

SDK availability & API documentation ↗
equity-actions.response.json
// Illustrative response · facing a bet
{
  "ok": true,
  "hand": "AsAh",
  "equity": 0.84,
  "facingBet": true,
  "output": {
    "fold": 0.02,
    "call": 0.26,
    "raise": 0.72
  }
}
Example values only. Full responses also include population targets and sizing weights.
KNOW WHAT YOU’RE BUILDING WITH

Designed for integration.
Grounded in the details.

Hand-aware, with a defined benchmark.

Equity is measured against a uniform heads-up opponent. It is not opponent-specific or multiway equity, and the policy is not a guarantee of optimal or profitable play.

Your range, when you have one.

Supply the actor’s candidate-range weights, or use a uniform board-compatible population. The endpoint does not infer or update ranges from action history.

Probabilities you can build around.

The API returns a distribution rather than selecting an action. Your application owns sampling, game rules, and what happens next.

WHILE WE DEAL THE NEXT HAND

Meet us at the table.

Public API access is on the way. Play Blitzfold in the meantime.

Explore Blitzfold

Building something with poker? Let’s talk about it ↗