Hotshot-XL is an open-source AI text-to-GIF model trained to work alongside Stable Diffusion XL: describe a scene and it generates a short animated GIF using any SDXL checkpoint, including your own fine-tunes and LoRAs. Released under the Apache-2.0 license, it has about 1,100 stars on GitHub and was last updated in 2024.
Text-to-video models are heavy and hard to personalize; Hotshot-XL took a clever shortcut. Instead of learning video from scratch, it adds temporal generation on top of the SDXL image ecosystem: the model was trained to produce one-second GIFs at 8 frames per second, across multiple aspect ratios, using images rather than scarce video data for supervision. Because generation rides on SDXL, any fine-tuned checkpoint or LoRA you already use for stills can drive your GIFs, which makes personalized characters and styles straightforward. It is also compatible with SDXL ControlNet, so composition and layout can be guided structurally frame by frame. For creators, the workflow slots into existing Stable Diffusion tooling: pick a model, write a prompt, and get a looping animation in the same style as your images.
Hotshot-XL is free and open source under the Apache-2.0 license. The model and code are available on GitHub and model hubs at no cost; you run generation on your own GPU.
Hotshot-XL is for Stable Diffusion creators, meme and sticker makers, and developers who want short, stylized animations generated in the exact look of their existing SDXL models and LoRAs.