0

Hanabi

Frontier

Hanabi game

Domain
rl-env
License
mit
Published
Dec 2025

Cite

Notes

Only stored in your browser.

Attribution

Leaderboard scores
prime-hub
Attribution policy →

Top score 12.93 by Grok 4 Fast - 12 models reporting (5 frontier)

Score history

8
03.757.511.2515Apr 25Jun 25Aug 25Oct 25Dec 25GPT-4.1 MiniGrok 4 Fast

Top models

12
HanabiBar chart with 12 bars. Highest value: GPT-5.2 at 12.9.
12 models

Related tools

1
View all

Implementations, trainers, datasets and scaffolds linked to this eval.

FAQ

What is Hanabi?
Hanabi game
What is the current top score on Hanabi?
The top reported score is 12.93 by Grok 4 Fast, across 12 models reporting (5 from frontier labs).
How can a model improve its Hanabi score?
Tools linked to Hanabi on Sophon include Hanabi RL Env (Community) - RL environments, datasets, and scaffolds that target this eval.
What license is Hanabi under?
Hanabi is available under mit.