Recraft Image Model - First AI with high-quality images in user-uploaded style

Hi Product Hunt community! My name is Anna Veronika and I’m the founder of Recraft. Right now Recraft provides free image generation for everyone, and allows commercial use of the generated images. Today I’m excited to announce the launch of our new image foundation model, Recraft 20B. This will provide our users with: - Better Style Control: You can now create a style from an image, or create a set of images in a consistent style. - Improved Anatomy: Our new model ensures improved anatomical correctness, paying attention to details like accurate human figures and correct fingers. - Enhanced Prompt Quality: Experience a smoother and more intuitive creative process as our model generates images that better align with user-written descriptions, ensuring even more impressive results. - Improved Details: small details like people's skin or objects in the background are now generated much more accurately We've come a long way in the past 8 months, and we can't wait for you to experience the future of graphic creation with Recraft. Visit https://www.recraft.ai/ to get started. Cheers to the next level of creative expression! 🚀

The icon generator looks impressive. I need to give Recraft a try and give you some feedback. I find that AI people often look "shiny". Does Recraft have the same challenge as well? Congrats on the amazing launch!

I really enjoyed using your design tool. I was pleasantly surprised at the generation quality as I was not expecting them to be so polished. I uploaded some pictures myself and performed the same actions and they looked brilliant! I am looking forward to the ability to upload SVGs. Also, is it already possible to make generative edits on text? If so, they were not obvious to me how one could do that. Congratulations on your launch! I will be tuned in to your progress

Want to create a set of images? Here's how easy it is now: 1. Click the 'set tool' in the toolbar to begin. 2. Click on the canvas to place a set frame. 3. Select a style you like. You can upload your own, choose a palette, or adjust settings. 4. Input up to 6 prompts and press 'Recraft' to start the generation. 5. Download the images individually by clicking on each and choosing 'export' in the top right corner. 6. To regenerate specific images, click on them, update the prompt, and recraft for a new style. 7. For individual edits, simply drag an image out of the frame and edit as needed. Give it a try and let us know what you think! Your feedback helps us improve.

Congrats on the launch!

thanks for everything, the best ai art generator mean vector art wtf it's f amazing, anna thanks a lot hope it will be free forever, i told hundreds of people about this

Congratulations on your new product launch! This is a product that we must have... this is the next level of creative expression

Good job team, best of luck today 🦄♥️

The new model looks like a big improvement, well done! Especially love how one can create a set of images in the same style. Wonder if this functionality could be extended to preserve e.g. the caracter (e.g. T-Rex driving a car, T-Rex reading a book, etc. with T-Rex character being the same throughout the series) or the setting (e.g. park in the snowy weather, park in the sunny weather with park being recognizably the same)?

I logged in, haven't used the app but one suggestion is the floating text on the input box has an background colour of white rather than the gray colour in the background. Will share my feedback after using. Congrats on the launch!