Google DeepMind's Four Gemini 4 Argon Agents Beat Minecraft's Ender Dragon Without Human Intervention

By AI123

Google DeepMind has published a paper describing how four Gemini 4 Argon agents beat the Ender Dragon in Minecraft without human intervention. The agents were placed in a randomized survival sandbox world with the goal of defeating the Ender Dragon, but were given no preset roles, strategies or game-specific tools.

Each agent controlled a player through the open-source Mineflayer library and coordinated via in-game chat and shared memory. During the run, the agents split up tasks on their own: one made tools, another gathered food and supplies, another lit a key portal, and another built Enderman shelters and landed the final blow. They also warned allies about threats, shared coordinates and tips, and recovered items from teammates who had died.

The whole process took 27.23 hours. According to the team, the four agents wrote and ran 2,783 programs, exchanged 229 messages and died 605 times along the way, using 1.35 billion input tokens. The final cost was $810 at Gemini's list price, or $1,620 at the standard price. A fifth agent, called the Observer, monitored the session as an independent witness to check for cheating and to publish live and summary reports.

The team said that, to their knowledge, this is the first end-to-end autonomous completion of Minecraft by AI agents, achieved using only the standard Mineflayer library without pre-scouted seeds or other game-specific tools written by humans. Winning required the agents to maintain a coherent plan across thousands of dependent steps for more than a day.

Explore topicsAgentGeminiGoogleModels

Source

View all