# 1X World Model

Models and data from the 1X World Model Challenge, which asks a model to predict the first-person camera view of 1X's EVE humanoid robot from its past frames and actions. The release includes small Genie-style baselines and a tokenizer; it is a research benchmark rather than an interactive app.

Organization: 1X Technologies. Released: June 2024. Kind: Robotics.

## What you control

- Past camera frames from the robot
- The robot's action data for each step

## License

Weights: Available (https://huggingface.co/1x-technologies/GENIE_138M). Declared license: Apache-2.0.
The weights' model card declares Apache-2.0, a permissive license that allows commercial use. Check what the model was built on, because a base model's terms can still apply.

## Real-time

No real-time claim.

## How to run it

1. Install the repository's dependencies and download the challenge data.
2. Generate frames with genie/generate.py and the 1x-technologies/GENIE_138M checkpoint, then view them with visualize.py.
3. Score the results with genie/evaluate.py.

## Where to try

No hosted destination verified.

## Sources

- [Project source](https://github.com/1x-technologies/1xgpt)
- [Weights and model card](https://huggingface.co/1x-technologies/GENIE_138M)

## Questions

### Is 1X World Model open source?

1X World Model publishes downloadable weights under Apache-2.0. The declared license is permissive.

### Can I use 1X World Model commercially?

The weights' model card declares Apache-2.0, a permissive license that allows commercial use. Check what the model was built on, because a base model's terms can still apply.

### Does 1X World Model run in real time?

The cited sources do not claim real-time operation.

### Where can I try 1X World Model?

No hosted version was verified; run the published weights yourself.

### What does 1X World Model take as input?

You give it past camera frames from the robot and the robot's action data for each step.

### How do I run 1X World Model?

Following the official instructions: Install the repository's dependencies and download the challenge data. Generate frames with genie/generate.py and the 1x-technologies/GENIE_138M checkpoint, then view them with visualize.py. Score the results with genie/evaluate.py.

### What are alternatives to 1X World Model?

Similar models in this directory include Ctrl-World (open weights, permissive license), EnerVerse-AC (open weights, non-commercial) and ABot-World (open weights, permissive license).

Checked 2026-10-23. Corrections: https://github.com/windflow-labs/world-model-directory