README.md

---
license: apache-2.0
pipeline_tag: image-to-video
language:
- en
base_model:
- Wan-AI/Wan2.2-TI2V-5B
- Wan-AI/Wan2.1-T2V-1.3B
tags:
- video-generation
- multi-player
- world-model
---

# Multics

Multics enables scalable multi-player video generation through hierarchical communication.

[Project page](https://karine-huang.github.io/Multics/) · [Code and instructions](https://github.com/TencentARC/Multics)

This repository contains the complete 4-player Multics DiT checkpoint, its configuration, the Wan2.2 VAE, the UMT5 text encoder, and tokenizer. The same checkpoint supports 6-, 8-, and 10-player generation without retraining.

From the Multics code repository, download all inference weights with:

```bash
hf download TencentARC/Multics --local-dir models
```

Then generate the first 4-player example:

```bash
python inference.py --config configs/inference_demo.yaml \
  --checkpoint models/multics/4players/checkpoints/model.ckpt \
  --hparams models/multics/4players/hparams.yaml \
  --case 1
```

See the code repository for installation, example inputs, and scalability commands.

The weights are distributed under the Apache 2.0 license. See [LICENSE](LICENSE) and [THIRD_PARTY.md](THIRD_PARTY.md) for component attribution.