Show HN: Lance – Image/video generation and understanding in one model

Show HN: Lance – image/video generation and understanding in one model

ByteDance released Lance as a research project for image and video generation and understanding in one model; the RSS snippet states 3B active parameters, fewer than 128 GPUs used for training, and links to a homepage, arXiv paper, and Hugging Face model, while the post does not disclose benchmark results or licensing terms.

Why it matters: ByteDance’s Lance puts image/video generation and understanding in one model, with 3B active parameters and <128 GPUs for training. HKR-H/K/R all pass, but benchmarks, license details, and real outputs are not disclosed, keeping it below P1.