Directory
Data from GitHub, checked 2026-10-16. Add yours.
Mindcraft LLM agent
LLM-driven Minecraft bot built on Mineflayer: chats, gathers, crafts and builds from text goals. Supports multi-agent collaboration.
- License
- MIT
- Models
- OpenAI, Anthropic, Gemini, Ollama and many other APIs
- GitHub stars
- 5,895
- Last activity
- 2026-10-03
Voyager LLM agent (research)
Open-ended embodied agent that writes and stores reusable skills as code, from the MineDojo team.
- License
- MIT
- Models
- OpenAI GPT-4 / GPT-3.5 (per paper)
- GitHub stars
- 7,268
- Last activity
- 2024-04-03
Mineflayer Bot framework
High-level JavaScript API for creating Minecraft bots. The foundation under Mindcraft and Voyager.
- License
- MIT
- Models
- N/A (library, bring your own AI)
- GitHub stars
- 7,585
- Last activity
- 2026-10-11
mineflayer-pathfinder Bot plugin
Pathfinding plugin that lets Mineflayer bots move from A to B.
- License
- MIT
- Models
- N/A
- GitHub stars
- 313
- Last activity
- 2026-10-06
MC-Bench Benchmark
Human-voted benchmark where LLMs write code to build prompted structures; voters compare builds blind. Site: mcbench.ai.
- License
- MIT
- Models
- Models from Anthropic, OpenAI, Google, Alibaba and others
- GitHub stars
- 168
- Last activity
- 2024-12-01
MineStudio Agent dev toolkit
Streamlined package for Minecraft AI agent development: simulator, data, models, training and inference.
- License
- MIT
- Models
- Bundled policy models (e.g. VPT-style) and custom
- GitHub stars
- 414
- Last activity
- 2026-09-28
MineDojo Research framework
Simulator and internet-scale knowledge base for building open-ended embodied agents.
- License
- MIT
- Models
- N/A (environment)
- GitHub stars
- 2,271
- Last activity
- 2024-03-18
Video PreTraining (VPT) Model / research
OpenAI's agent that learned to play Minecraft by watching unlabeled online videos.
- License
- MIT
- Models
- Own VPT models
- GitHub stars
- 1,751
- Last activity
- 2025-09-03
MineRL RL environment
Python package for the MineRL sample-efficient reinforcement learning competition.
- License
- Custom (see repo)
- Models
- N/A (reinforcement learning)
- GitHub stars
- 980
- Last activity
- 2025-01-22