Dalle3 is Free! showcase and tips to use with SDXL

#dalle #dalle3 #aiimagegeneration #aiimages #stablediffusion #a1111 #
00:00:00 sample Dalle 3 images
showing prompts used and example images
00:00:39 how to use Dalle 3 from Bing chat
00:01:31 improving Dalle output using Stable diffusion
improve faces and realism of the Dalle 3 images using stable diffusion in A1111
This short video will display sample images generated using Dalle 3 free Bing chat, which shows how complicated, accurate, and beautiful its results can be.
Currently, Bing chat allows us to generate these images for free, which Based on my tests are very simple to generate and alot better than SDXL or Midjourney in terms of composition, accuracy and adherence to the prompt.
We will also show some notes on how to improve images that contain people, in some cases using SDXL refiner model in img2img.
100% of the images created in Dall.E 3 are just perfect
It understands complex sentences with high level of details more than any other image generation software.
Images generated can be further refined in stable diffusion SDXL refiner model.
Single character generation in SDXL is more flexible and realistic than Dalle 3

[@DavidSeguraIA]

Please share your prompts for the female dalle 3 images, please

[@AI-HowTo]

Generate a stunning and highly detailed digital artwork depicting a futuristic robotic woman composed of gleaming liquid metal, her intricate wiring visible through translucent layers of her body. This robotic woman should exude an aura of advanced technology and sophistication, with dynamic lighting playing off the curves and contours of her liquid metal form. The scene should be set in a futuristic cityscape at dusk, with neon lights reflecting off her reflective surface, creating a striking contrast between her sleek, cybernetic appearance and the bustling urban environment. The attention to detail should be so exquisite that viewers can appreciate the intricate craftsmanship that went into creating this robotic marvel, down to the smallest wire and liquid metal ripple. The overall mood of the image should be both awe-inspiring and slightly eerie, evoking a sense of wonder at the intersection of technology and humanity

[@AI-HowTo]

all images prompts are included under the image, only the third was shortened, because I asked ChatGPT to give me a more detailed prompt...i think only first three lines are enough to generate similar image though,,,, but ChatGPT can be more creative to give more details in less time

[@user-kk2ve1un4u]

actually you are using the generated original img2img in the stable diffusion A1111 with a denoising strength 0.4 and a sdxl refiner model which gave more realistic than the original . It is defficult to compare between both dalle3 and SDXL webui by using text2img since both don't have the same seed and settings .

[@AI-HowTo]

I tried like 100 images + with Dalle 3 and SDXL, Dall3 results was actually better in all of them, the level of details, accuracy, composition, adherence to the prompt, I didn't expect Dalle 3 to be this good actually, SDXL results were only better when generating Single characters and portraits of people... I like SDXL and its freedom and flexibility, but honestly, Dalle 3 is like another level and generation of complex images is simpler with less prompting.

[@AI-HowTo]

when it comes to Characters, refining Dalle 3 output using SDXL produces better results, Stable diffusion in paint features are essential for people to make faces look better and introduce more changes to the output....I think in general, that using multiple tools is essential in this AI Sector, one alone simply doesn't work well.