Antim Labs wordmark

TEACHING MACHINES THROUGH GAMES

Antim Labs is a research and product company pioneering game-based training for intelligent systems. We build interactive game environments and capture human gameplay with reasoning traces to enable the next frontier of AI.

Games compress the messy complexity of the real world into loops of perception, reasoning, and action.

Antim Labs builds games where machines learn the way humans do: by playing, exploring under constraints, and making high-stakes decisions at low cost. Our environments and gameplay logs generate the interactive data needed to train world models for open-ended learning, spatial intelligence, and embodied AI.

Game Environments

Endless
Flappy Bird

Flappy Bird

LLM-friendly version of the viral mobile hit with text-scaffolded physics. Moves are turn-based—agent receives current position, velocity, and upcoming pipe gaps, then decides TAP or do nothing. Physics constants are explicit so the model can project trajectories ahead.

Spatial
Snake

Snake

Classic Snake on an 8×8 grid, but the agent plans full paths to each apple rather than single moves. The challenge: reasoning about when body segments will clear before the head arrives. An HP system discourages guessing—reliable pathfinding beats trial-and-error.

Probabilistic
Blackjack

Blackjack

The classic card game, text-scaffolded for LLMs. Agent sees its hand total and the dealer's face-up card, then chooses HIT or STICK. Dealer plays by fixed rules (hits on 16 or below). Pure probabilistic reasoning—when to push your luck vs. stand pat.

Route Optimization
Taxi

Taxi

A 5×5 city grid with walls and four pickup/dropoff spots. Agent drives to the passenger, picks them up, navigates around walls, and drops them at the destination. Multi-step logistics where planning the full route matters.

Long Horizon
Catan

Catan

1v1 Settlers of Catan, first to 7 Victory Points wins. Agent manages resources, trades, builds settlements and roads. Long-horizon strategy where early positioning shapes late-game options.

view all

Research Preview

From Game Replays to Generalization

We trained a multi-task RL model across nine interactive game environments to study how skills learned in games transfer out of distribution. This work is grounded in our thesis that minds are scaffolded by the environments they act in, that intelligence lies in the loop between an agent and its world. We created the environments inspired by classic arcade games and traditional RL benchmarks, training separate single-game experts plus a shared multi-task agent over all of them. We then evaluated these agents both in-game and on a diverse set of downstream math, coding, and reasoning benchmarks.

Full report coming soon.