Play against HexZero, HexGPT or Random.

Tiny play models, based on AlphaZero and GPT architectures, trained on MacBook in 2 hours.

Player 1 (Blue, top → bottom)

Player 2 (Orange, left → right)

Blue's turn