Skip to content

Factorio学习环境

An open-ended framework for evaluating LLM planning

Visit official site

What Factorio Learning Environment is

The framework is an open-ended evaluation environment built on the game Factorio for testing LLM planning, program synthesis and resource optimisation.

What you can do with it

  • Run open-ended challenges from basic automation to complex factories
  • Choose between Lab-play and Open-play protocols

What to watch out for

  • Evaluation is built around one game environment
  • The Python API is the stated interaction path

Pros & cons

✓ What we like

  • Open source and free to use
  • Production scores and milestones measure planning

! What to watch out for

  • Evaluation is built around one game environment
  • The Python API is the stated interaction path

FAQ

What does it evaluate?

Long-term planning, program synthesis and resource optimisation in LLMs.

How many structured tasks?

The material states 24 structured tasks in Lab-play.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review