Lance: Unified Multimodal Modeling by Multi-Task Synergy

Fengyi Fu*, Mengqi Huang*,✉, Shaojin Wu*, Yunsheng Jiang*, Yufei Huo, Jianzhu Guo✉,§ Hao Li, Yinghang Song, Fei Ding, Qian He, Zheren Fu, Zhendong Mao, Yongdong Zhang
ByteDance
* Equal contribution    ✉ Corresponding authors    § Project lead

🌟 Highlights

Lance is a 3B native unified multimodal model that supports image and video understanding, generation, and editing within a single framework.

• Efficient at 3B scale. With only 3B active parameters, Lance delivers strong performance across image generation, image editing, and video generation benchmarks.
• Trained from scratch. Lance is built with a staged multi-task recipe and trained entirely from scratch (except for the ViT and VAE encoders; the transformer backbone is trained entirely from scratch) within a 128-A100-GPU budget.

We are actively updating and improving this repository. If you find any bugs or have suggestions, please feel free to open an issue or submit a pull request (PR) 💖.

🎨 Demo

Text-to-Video

Video Editing

Multi-turn Consistency Editing

Intelligent Video Generation

Video Understanding

Text-to-Image Generation

Image Editing

Image Understanding

🚀 Installation

Recommended Environment

• Software: Python 3.10+, CUDA 12.4+ (required)
• Hardware: A GPU with at least 40GB VRAM is required for inference

Installation Steps

bash ./setup_env.sh

Download Model Weights

Please download all necessary model checkpoints from Lance-3B on Hugging Face and place them in the downloads/ directory.

📚 Usage

Inference

We provide a unified command-line interface for all generation / editing / understanding tasks:

Option 1: Configure and Run the Unified Script

bash inference_lance.sh

• Before running, please configure the inference parameters at the top of inference_lance.sh.
• Supported tasks: t2i, t2v, image_edit, video_edit, x2t_image, and x2t_video. You can modify TASK_DEFAULT_CONFIGS in inference_lance.py to customize the default data samples for each task.
• Note: For all tasks, we recommend following the prompt format used in the provided examples when writing input prompts, as this typically leads to better generation quality.

Option 2: Configure and Run the Unified Script

We provide task-specific one-click commands for different generation, editing, and understanding tasks.

Text-to-Video Generation

bash inference_lance.sh \
  --TASK_NAME t2v \
  --MODEL_PATH downloads/Lance_3B_Video \
  --RESOLUTION video_480p \
  --NUM_FRAMES 121 \
  --VIDEO_HEIGHT 480 \
  --VIDEO_WIDTH 848 \
  --SAVE_PATH_GEN results/t2v

Text-to-Image Generation

bash inference_lance.sh \
  --TASK_NAME t2i \
  --MODEL_PATH downloads/Lance_3B \
  --RESOLUTION image_768res \
  --VIDEO_HEIGHT 768 \
  --VIDEO_WIDTH 768 \
  --SAVE_PATH_GEN results/t2i

Video Editing

bash inference_lance.sh \
  --TASK_NAME video_edit \
  --MODEL_PATH downloads/Lance_3B_Video \
  --RESOLUTION video_480p \
  --SAVE_PATH_GEN results/video_edit

Image Editing

bash inference_lance.sh \
  --TASK_NAME image_edit \
  --MODEL_PATH downloads/Lance_3B \
  --RESOLUTION image_768res \
  --SAVE_PATH_GEN results/image_edit

Video Understanding

bash inference_lance.sh \
  --TASK_NAME x2t_video \
  --MODEL_PATH downloads/Lance_3B_Video \
  --RESOLUTION video_480p \
  --NUM_FRAMES 50 \
  --SAVE_PATH_GEN results/x2t_video

Image Understanding

bash inference_lance.sh \
  --TASK_NAME x2t_image \
  --MODEL_PATH downloads/Lance_3B \
  --RESOLUTION image_768res \
  --SAVE_PATH_GEN results/x2t_image

Available Tasks

For understanding examples:

• config/examples/x2t_image_example.json: image understanding examples for visual question answering and image-based reasoning.
• config/examples/x2t_video_example.json: video understanding examples for video question answering and video captioning.

Parameters

You can configure the following hyperparameters at the top of the inference_lance.sh script:

Gradio

python lance_gradio_t2v_v2t.py --gpus 0 --server-port 7860

Benchmarks

† indicates methods that use LLM rewriters for prompt rewriting before generation.

Running Benchmarks

Ready-to-run benchmark scripts are provided under benchmarks/:

📄 License

Copyright 2025 Bytedance Ltd. and/or its affiliates.

🙏 Acknowledgements

We would like to thank the contributors of BAGEL, Qwen2.5-VL-3B-Instruct, and Wan2.2 for their open research and contributions.

💖 Citation

If you find Lance useful for your project or research, welcome to 🌟 this repo and cite our work using the following BibTeX:

@misc{fu2026lanceunifiedmultimodalmodeling,
      title         = {Lance: Unified Multimodal Modeling by Multi-Task Synergy},
      author        = {Fengyi Fu and Mengqi Huang and Shaojin Wu and Yunsheng Jiang and Yufei Huo and Hao Li and Yinghang Song and Fei Ding and Jianzhu Guo and Qian He and Zheren Fu and Zhendong Mao and Yongdong Zhang},
      year          = {2026},
      eprint        = {2605.18678},
      archivePrefix = {arXiv},
      primaryClass  = {cs.CV},
      url           = {https://arxiv.org/abs/2605.18678},
}

📞 Contact

For questions, issues, or collaborations, please contact Mengqi Huang and Jianzhu Guo.