# AI agent advancements in Minecraft automation > Developing situation record from CLSTR: https://clstr.news/situations/ai-agent-advancements-in-minecraft-automation > Updated: 2026-10-17T03:13:40.000Z. Sources: 7. Developments: 2. Developments in AI agent capabilities have been demonstrated through complex task completion within the game Minecraft. Initially, engineer Ronak Malde utilized a dual-model strategy to defeat the Ender Dragon in under nine minutes. This approach combined OpenAI’s ‘GPT-6 Astra’ as a high-level planner with TypeSafe AI’s ‘Jev’ acting as a rapid decision-maker to execute immediate actions. This architecture allowed for continuous movement while the larger model processed strategic steps. Subsequently, Google DeepMind demonstrated a different autonomous approach using four agents powered by the Gemini 4 Argon model. These agents completed a full playthrough in approximately 27 hours without human intervention. Unlike previous methods, these agents did not rely on visual interpretation but instead used the Mineflayer JavaScript library to write and execute code for navigation and combat. The group coordinated through in-game chat and shared notes, generating over 63,000 lines of code to overcome challenges such as multiple deaths and server restarts. ## Timeline ### 2026-10-17: Google DeepMind AI agents complete Minecraft autonomously Four Google DeepMind AI agents using Gemini 4 Argon autonomously completed a Minecraft survival match, defeating the Ender Dragon in 27 hours through self-written code and coordination. 3 sources. https://clstr.news/cluster/google-deepmind-ai-agents-complete-minecraft-autonomously ### 2026-09-20: AI agents defeat Minecraft Ender Dragon using dual-model strategy An AI agent combining OpenAI’s Astra and TypeSafe AI’s Jev defeated the Minecraft Ender Dragon in 8 minutes and 43 seconds using a dual-system planning and rapid-decision architecture. 4 sources. https://clstr.news/cluster/ai-agents-defeat-minecraft-ender-dragon-using-dual-model-strategy --- Cite as: AI agent advancements in Minecraft automation. CLSTR, https://clstr.news/situations/ai-agent-advancements-in-minecraft-automation