# Lance

> Open-source, self-hosted unified multimodal model from ByteDance: 3B active parameters covering text-to-image, video synthesis, media editing, and visual Q&A.

- **Category:** AI
- **Licence:** Apache-2.0
- **Open source:** oss
- **Languages:** Python
- **Platforms:** Python, Build from Source
- **Tags:** Multi modal, Image Generation, Video Editing, Text to image, Diffusion, Text To Video, Video Generation, Vision Language Model, Image To Video, Image Editing, Video Understanding, Image Understanding, Unified Multimodal Models

## Description

Lance, from ByteDance's Intelligent Creation Lab, is a native unified multimodal model that handles understanding, generation, and editing of both stills and clips within a single 3B-active-parameter network. Its inference code exposes seven tasks: text-to-image, text-to-video, image-to-video, image editing, video editing, and image or video understanding for captions and visual Q&A, while the galleries demonstrate chained multi-round edits and reasoning clips like a solved maze. T2V and I2V runs can first send the prompt to any OpenAI-compatible endpoint for rewriting, and one local Gradio front end (port 7860) covers every task. Self-hosting means cloning the repo, building a CUDA 12.4+ Python environment, downloading the video-capable or image-only checkpoint from Hugging Face into the downloads folder, then calling the supplied shell script; vLLM-Omni documents a serving recipe, and released scripts fine-tune on local parquet data. The maintainers call it a research artifact, not a polished product: training stayed under a 128 A100 GPU budget, topping out at 768x768 stills and 480p 12 fps footage, so output quality fluctuates and no container images or tagged releases exist.

## Links

- [Homepage](https://github.com/bytedance/lance)
- [Website](https://lance-project.github.io)
- [Technical report on arXiv](http://arxiv.org/abs/2605.18678)

---
Source: https://selfhost.directory/project/lance