Newsletter
Join the Community
Subscribe to our newsletter for the latest news and updates
Unleash cinematic AI video generation locally with Sulphur 2—the unrestricted, open-source 9B parameter model built for ultimate creative freedom.
Traffic, search & AI signals for www.sulphur2.app.
Third-party traffic estimate · Updated Aug 9, 2026
Submit your own product to reach creators and founders looking for the next tool to try.

A pact network that helps founders prepare for their Product Hunt by finding support from fellow founders.

AI sales agent that automates LinkedIn outreach to find prospects, send messages, and book meetings.

Free online audio to text converter that turns MP3 and 45+ audio formats into 99-language transcripts with timestamps and speaker labels.

A free browser-based 3D ritual where you break a virtual willow branch and receive an AI-written consequence for your wish.

Weekly startup launch platform where founders showcase new products, get community votes, and grow faster.

Interactive YouTube embeds with polls, quizzes & lead form
Sulphur 2 is an open-source, uncensored 9B-parameter video generation model fine-tuned from LTX architecture, designed for local execution via ComfyUI to produce cinematic-quality videos from text and image prompts without cloud dependency or content moderation.
Sulphur 2 is a 9-billion parameter text-to-video and image-to-video model built on the LTX architecture and optimized for local inference. It takes written descriptions or static images as input and generates video clips with natural motion, smooth camera movements, and temporal coherence. The model runs entirely on NVIDIA GPUs (24GB+ VRAM recommended) within the ComfyUI environment and is released under the MIT open-source license by SulphurAI. Download weights or view showcases.
huggingface-cli download SulphurAI/Sulphur-2-base to get the 9B parameter checkpoint (placed in ./models/checkpoints).workflow.json into ComfyUI, connect your input (text or image), and click Queue Prompt. Sulphur 2 runs entirely on your NVIDIA GPU with no external service calls.Pros
Cons
Yes, Sulphur 2 is completely free and open source under the MIT license. You only pay for your own hardware and electricity.
A computer with an NVIDIA GPU supporting CUDA. Minimum 12–16 GB VRAM for quantized weights (reduced quality); 24 GB or more recommended for full 9B parameter native resolution. Integrated graphics and Apple Silicon are not supported.
No, currently Sulphur 2 does not support Apple Silicon (M1/M2/M3) or any non-NVIDIA GPU. Intel/AMD graphics cards are also not supported.
Install ComfyUI, download the model weights from Hugging Face, and load the official workflow file. Detailed steps are provided on the Sulphur 2 website.
Yes, Sulphur 2 natively supports both text-to-video and image-to-video generation within the same model.