jdopensource / JoyAI-Echo

Abstract

Long video generation still suffers from error accumulation, weak temporal coherence, and prohibitive latency, limiting its applicability to interactive scenarios. We present JoyAI-Echo , a framework that breaks these barriers through four key advances. Central to its performance, a cross-modal audio-visual memory bank preserves character appearance and voice timbre consistently over five-minute videos, while a post-training pipeline combines memory-based reinforcement learning with distribution matching distillation for a 7.5× speedup to substantially boost visual quality and alignment. Empowered by these two components, JoyAI-Echo decisively outperforms HappyOyster (directing mode) on long-form generation and even surpasses the short-video specialist Wan 2.6 on human-centric tasks. Beyond raw generation quality, an interactive agent enables real-time user editing through conversational instructions, and a lightweight super-resolution module maintains high definition under streaming latency, further elevating the overall experience and delivering instantly editable, conversation-speed video creation. For the first time, JoyAI-Echo simultaneously achieves long-range cross-modal consistency, real-time inference for minute-long video, conversational interactivity, and high-resolution output — without compromise, inaugurating a new era of interactive video generation. Codes and weights will be open-sourced.

Highlights

• 🎞️ Minute-level multi-shot stories : generate a sequence of coherent shots from one prompt JSON.
• ⚡ DMD-distilled few-step inference : ~7.5x faster than the original pipeline.
• 🔊 Joint audio-video generation : one pipeline produces synchronized video and audio.
• 🧠 Paired cross-modal memory bank : conditions each new shot on prior visual identity and voice context for story-level consistency.

Demo Gallery

Explore long-form and short-form JoyAI-Echo cases on the Project Page . 🍿

Results

Reported Scale

Human Evaluation

GSB user study on long- and short-video generation. The numbers denote the percentage of user preferences.

Quickstart

1. Clone

Get the Repo at first!

git clone https://github.com/jd-opensource/JoyAI-Echo.git
cd JoyAI-Echo

2. Create the environment

The reference environment is Python 3.11 + PyTorch 2.8 + CUDA 12.8 .

With conda:

conda env create -f environment.yml
conda activate echo-long

With uv :

uv venv --python 3.11 .venv
source .venv/bin/activate
uv pip install --extra-index-url https://download.pytorch.org/whl/cu128 -r requirements.txt

ffmpeg must be available on PATH for shot concatenation. The conda recipe includes it. If you use uv , install it with your system package manager:

sudo apt install ffmpeg
# macOS:
brew install ffmpeg

3. Download checkpoint

Download the JoyAI-Echo release checkpoint and Gemma text encoder:

Place them under checkpoints/ :

checkpoints/
+-- echo-longvideo-release.safetensors
`-- gemma-3-12b/

4. Write a story prompt

Create a JSON file under prompts/ .

Each string is one complete shot description. A single prompt creates a single shot. Multiple prompts create a multi-shot story conditioned through the paired audio-video memory bank.

5. Run inference

python inference.py

This loads the model once and processes all prompt files under prompts/ .

💡 Note : The inference pipeline is optimized to run on lower-VRAM GPUs. Peak GPU usage is around 46–50 GB , at the cost of slightly longer per-shot inference time.

Outputs are written to:

inference_result/outputs/<prompt-name>/inference_<timestamp>/

Configuration

All inference parameters are managed in configs/inference.yaml . The file is organized into sections:

Override via CLI

Any YAML parameter can be overridden from the command line:

python inference.py --seed 42 --num-frames 121 --video-height 480 --video-width 832

Use a custom config file:

python inference.py --config configs/my_experiment.yaml

The Python entrypoint exposes the full configuration surface:

python inference.py --help

Hardware

Peak GPU usage is around 46–50 GB for the default 25 fps x 241 frames x 1280 x 736 setting, so a single H100/A100-class (80 GB) or 48 GB GPU is sufficient.

For smaller GPUs, reduce resolution/frames:

python inference.py --num-frames 121 --video-height 480 --video-width 832

TODO List

• Release inference code
• Release model checkpoints
• Add prompt examples
• Release Director Agent

Links

• Project page: https://echo-team-joy-future-academy-jd.github.io/Echo-LongVideo-Page/
• Repository: https://github.com/jd-opensource/JoyAI-Echo
• huggingface: https://huggingface.co/jdopensource/JoyAI-Echo

Acknowledgements

We gratefully acknowledge the open-source projects this work builds upon — in particular LTX2.3 for the base video generator and Gemma for the text encoder. Thanks to the broader research community whose contributions made this release possible.

Citation

If JoyAI-Echo helps your research or products, please cite:

@techreport{echo2026longvideo,
  title        = {JoyAI-Echo: Pushing the Frontier of Long Video Generation},
  author       = {{Echo Team @ Joy Future Academy, JD}},
  institution  = {Joy Future Academy, JD},
  year         = {2026},
  month        = {May}
}

License

This project is based on LTX-2 by Lightricks Ltd.

Portions of the original LTX-2 codebase have been modified by JD.com for academic and research purposes only. This project is not intended for commercial use. For commercial use of LTX-2 or its derivatives, please contact Lightricks Ltd.

All original copyright, license, patent, trademark, and attribution notices from LTX-2 are retained. This project remains subject to the LTX-2 Community License Agreement.