Gemini 4 Argon agents defeat Minecraft’s Ender Dragon in 27 hours

Four Google DeepMind Gemini 4 Argon AI agents autonomously completed Minecraft and defeated the Ender Dragon in 27 hours and 14 minutes, dying 605 times without human intervention, according to a research paper released Thursday. The agents wrote and executed 2,783 JavaScript programs, exchanged 229 messages, generated roughly 63,313 lines of JavaScript and developed 42 reusable helper functions while processing about 1.35 billion input tokens and 16 million output tokens across roughly 24,000 model calls. DeepMind said the run, which cost approximately $520 at introductory Gemini 4 Argon API pricing and $1,040 at standard pricing, is the first known autonomous end-to-end Minecraft completion by general-purpose AI agents without custom gaming tools or predetermined world configurations.

Gemini 4 Argon agents defeat Minecraft’s Ender Dragon in 27 hours Four autonomous AI agents defeated the Ender Dragon in 27 hours, overcoming 605 deaths without human intervention or predefined gaming strategies. Google https://cryptobriefing.com/markets/alphabet/ DeepMind researchers have demonstrated that four Gemini 4 Argon AI agents can autonomously complete Minecraft, defeating the game’s final boss after more than 27 hours of continuous collaboration. According to a research https://arxiv.org/pdf/2610.22319 paper https://arxiv.org/pdf/2610.22319 released Thursday, the agents wrote and executed 2,783 JavaScript programs, exchanged 229 messages and died 605 times before defeating the Ender Dragon. The experiment took place in a randomly generated Minecraft survival world on Easy difficulty. Researchers provided the agents with a single objective but did not assign roles, develop gaming strategies or intervene during the run. DeepMind described the result as the first known autonomous completion of Minecraft from beginning to end using general purpose AI agents without custom gaming tools or predetermined world configurations. Gemini agents coordinate and develop their own strategies Each agent controlled a Minecraft character through Mineflayer, an open source JavaScript library that allows programs to interact with the game. Rather than using a keyboard or interpreting gameplay visually, the agents wrote code to navigate the world, gather resources, craft equipment and fight enemies. Within the first ten minutes, they began dividing responsibilities. One gathered supplies, another focused on tools and exploration, while the others prepared for the game’s more challenging environments. The agents also maintained shared memory, recording locations, strategies and lessons learned from previous failures. During the experiment, they developed 42 reusable helper functions to automate activities such as movement, resource collection and survival. AI, tech, and the markets they move—in one daily briefing. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. One agent even calculated the location of a hidden stronghold by tracking the movement of two Eyes of Ender and determining where their paths intersected. The estimate came within 43 meters of the actual portal room. However, the team encountered significant difficulties after entering the End, where the dragon repeatedly killed its members. The breakthrough came when the agents identified ten crystals that continuously restored the dragon’s health. They developed a strategy using stacked dirt blocks to reach and destroy the crystals before finally defeating the boss. The entire process took 27 hours and 14 minutes, including six server restarts. Experiment highlights advances in autonomous AI agents Beyond completing Minecraft, the experiment tested whether general purpose AI systems could coordinate across thousands of dependent tasks without continuous human guidance. The four agents collectively generated approximately 63,313 lines of JavaScript and maintained a shared strategy over more than a day of activity. According to the researchers, the agents processed approximately 1.35 billion input tokens and generated 16 million output tokens across roughly 24,000 model calls. Based on Gemini 4 Argon’s introductory API pricing, the cost of completing the game was approximately $520, rising to $1,040 at standard pricing. The researchers argued that the experiment demonstrates capabilities relevant to complex software engineering and enterprise workflows, where AI systems must plan, coordinate and recover from failures over extended periods. However, the results were achieved in a controlled gaming environment using programmatic access to game information rather than visual interaction. The researchers also acknowledged that the experiment does not establish whether the same capabilities would transfer directly to real world environments, where tasks may be less predictable and mistakes more costly. Nevertheless, the findings highlight progress in AI systems capable of maintaining objectives, developing tools and collaborating autonomously across lengthy sequences of actions. Disclosure: This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our Editorial Policy https://cryptobriefing.com/editorial-policy/ .