DiffusionKit

Hello Product Hunt! I'm stoked to introduce my take on a generative AI SaaS. This is very much an MVP, but even at such an early stage there are some noteworthy differentiators here (when compared to, say Midjourney, DALLE, or whatever): • DiffusionKit uses replicate.com for its backend, which means at present you can choose between Stable Diffusion 2 or OpenJourney (a MidJourney model for Stable Diffusion) • DiffusionKit also offers faceswapping/deepfaking, which makes it particularly potent for concept art in the entertainment industry (this is my personal use case), album art, avatars, etc • Updates/new models will be added regularly. Dreambooth & inpainting are coming before the end of the year. • In the future, you will be able to use your own generative AI models (whether SD or something else) simply by adding API keys for Replicate, Banana, Paperspace, etc, and sell access to them for a markup. • Finally, pricing: nobody else is offering this, but on the months that you don't use it, you will not be charged at all. I arrived at this insight after realizing that after my initial curiosity about Midjourney, RunwayML, etc had waned, my usage dropped to nearly nothing, yet I was paying a full $20-30 on each platform. By reducing both the barrier to entry and the cost to remain a user to near-zero, I hope to make this a no-brainer for those who are curious, but not looking for a big commitment. Looking forward to seeing what you guys make, and hearing what you think of the app. Technical BTS: this MVP is built on Node + Vue + Nuxt + Bootstrap-Vue. Hosted on Netlify (static site + serverless API). Payment via Stripe Checkout. The MVP took me 10 days to build. After building many MVP's I'm stoked to have my Time-To-Launch down to under 2 weeks!