This release adds new games, bug fixes, and process + support changes.

Details coming soon!

Support and Process changes

• Mark TF algoithms as unmaintained
• Mark Go API as unmaintained
• Mark Rust API as unmaintained
• Remove R-NaD
• Upgrade version of pybind11 and use master branch (post-merge of smart-holder branch into master)
• Add support for Python 3.13

Games

• German Whist
• Einstein Wurfelt Nicht
• (m, n, k)-Game
• Hive
• Cribbage
• Abrupt PTTT

Algorithms

• EFR (Morrill et al. 2021, Efficient Deviation Types and Learning for Hindsight Ratonality in Extensive-Games)
• PopRL and IMPALA for repeated RPS (Lanctot et al. 2023, Population-based Evaluation in Repeated Rock-Paper-Scissors as a Benchmark for Multiagent Reinforcement Learning)
• SCO (Lanctot et al. '25, Soft Condorcet Optimization for Ranking of General Agents)
• Asynchronous MCTS (Schultz et al. Mastering Board Games by External and Internal Planning with Language Models)

Improvements and other additions

• Update Deep CFR to implement Policy
• UCI bot: run matches more quickly, canonical path, concurrency fixes
• Add StepVerbose to the Bot class
• Chess: expose IsRepetitionDraw
• Spades updates
• Expose NewInitialState(const std::string& str) to python
• Expose TurnBasedSimultaneousState and internal simultaneous game state to Python
• Breakthrough serialization: add player's turn info
• Add GetRandomCategoricalPolicy
• Rand-psAR action translation in Poker
• Blackjack: add ResampleFromInfoState
• Add is_concrete field to GameType to separate concrete games from game wrappers/transforms
• Core API: allow rewards to be returned at chance nodes
• StatCounter and sliding window averager for helpful compiling of results

Fixes

• Big/typo to IS-MCTS to work with perfect information games
• DQN target for illegal actions
• Chess960: castling logic
• Chess960: move to action
• Information set strings in PTTT
• AlphaZero: chance nodes, tensors with low dimension, json output
• Blackjack: information state tensor, observation tensor
• Max game length of TTT variants
• Exploitability usage example in sequence-form LP

Documentation

• Replace flat list of games with a table
• Improve Windows compatibility
• Add a link to OpenSpiel visual graph
• Add missing references to mean field games

Acknowledgments

Thanks to Google DeepMind for continued support of development and maintenance of OpenSpiel.

Thanks to all of our contributors:

• Core Team: https://github.com/deepmind/open_spiel/blob/master/docs/authors.md
• All Contributors: https://github.com/deepmind/open_spiel/graphs/contributors