MineDojo/Voyager

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

7160 stars · 700 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Voyager is an open-source LLM-powered embodied agent that plays Minecraft through GPT-4, using an automatic curriculum, an ever-growing skill library of executable code, and iterative prompting with environment feedback. It is the reference implementation of the Voyager research paper on open-ended embodied lifelong learning agents.

Use cases

• run an LLM agent that autonomously explores Minecraft
• research lifelong learning for embodied AI agents
• build agents that write and store executable skills as code
• reproduce the Voyager paper experiments
• benchmark LLM agents on open-ended exploration tasks
• study iterative prompting with environment feedback and self-verification

When to choose

• you want an open-ended LLM-driven agent in Minecraft
• you are researching embodied lifelong learning or skill libraries
• you want a code-generation-based agent without model fine-tuning

When to avoid

• you need a production agent for real-world robotics or non-Minecraft environments
• you cannot provide GPT-4 API access or a Minecraft instance
• you need a lightweight agent without heavy game and Node.js dependencies

Facets

library · maturity active

agent-framework llm-inference machine-learning game artificial-intelligence large-language-models reinforcement-learning gaming-tools python windows embodied-agents lifelong-learning minecraft gpt-4 skill-library open-ended-learning research-code ai-agents research nodejs linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="MineDojo/Voyager")