An uncensored video generation model based on LTX 2.3 supporting both t2v and i2v natively, as well as all of the other ltx 2.3 formats.
Downloads
52,477
Sulphur-2-Base-GGUF is the quantized GGUF distribution of SulphurAI's open-source video generation model Sulphur-2-Base — a heavily fine-tuned LTX 2.3 model with ~21B parameters, supporting both text-to-video and image-to-video generation for fully local inference via ComfyUI.
Sulphur-2-Base is developed by SulphurAI and published on Hugging Face. It is built as a deep fine-tune of the LTX 2.3 ecosystem, with a strong focus on delivering high-quality, unrestricted video generation entirely on local hardware — no cloud platform, no subscription required. The GGUF versions are community-packaged quantized builds (published by contributors such as vantagewithai and Abiray) specifically optimized for local inference tools like ComfyUI.
Sulphur-2-Base natively supports Text-to-Video (T2V) and Image-to-Video (I2V), along with all other LTX 2.3 generation formats. Community feedback highlights notable improvements in motion consistency — smoother camera movement, more realistic human animation, and stronger frame-to-frame continuity — all contributing to a noticeably cinematic output quality. The model also supports LoRA integration and deep workflow customization, making it a compelling choice for advanced ComfyUI users.
