0:47 WTF? 😲 Dall-E is still the best for most images i create and I've tested it against all the models including Recraft. It's true that each model has it's strengths and weakness, but no model is as good as Dall-E for prompt adherence and off training creativity.
Wow some real negative commenters. Most do not understand that different people want different things and different tools. Just because you are not a fan does not mean it is bad. There are many different jobs out there that require different tool based on different skillsets. And dropping a comment that you do not like it is rather pointless, we all have opinions, does not mean we need to all voice them. I like to be across ALL the tools available so I can recommend the right tool for the right person, so thank you for this video and all the others. 🙏🖤
Not bad, but in your video im not really seeing anything ground breaking that makes me go 'wow flux /SD or MJ cant do anything like that' When new Generative text to image AI tools come out, why do so many inflencers test them with yet another a cute cartoon cat, or cyber punk city scape, or a pretty woman portait ? -Stuff that we know all the AI modles to date can already do just fine, SD 1.5 can do those just fine. Youre all going far to easy on them with all this safe stuff! :D Give it 'a guy throwing a ball at a lamp post' or 'a robot breaking though a wall' or 'a woman laying down or doing a high kick' or hell, just a just basic train / bus interior that doesnt look like it was designed and built by smack heads
I think people are approaching it from 2 directions currently. Researchers who want to push the limits, but have less focus on UX and UX focused tools with less interest in pushing the limits. Really hard to do both right now.
Not much inspiring about a blinding canvas. The void is described as satin black, and you create from there -- not from bright light. What year is this? Great branding on the hat, Mr. Kamph! Thanks for all your work! Yellow is pretty great. Love.
When I see styles to chose I'm usually out. Means you can not do your own styles and therefor useless for serious work. Also this is not testing. This is a joke.
Test Recraft AI at go.recraft.ai/sebastiankamph - Use code SEBASTIAN12 for $12 off
Nice but I still prefer the control and ability to craft specialized workflows I have with comfy. No simple and pretty UI is gonna beat that
While I agree that Comfy is great, we're not limited in only using one tool 😊
I wasn't super impressed knowing it was a paid deal, but then that marketing example with the coffee was pretty F'ing fluid... that was kinda nuts.
Yeah, they got some cool features built in.
0:47
WTF? 😲 Dall-E is still the best for most images i create and I've tested it against all the models including Recraft.
It's true that each model has it's strengths and weakness, but no model is as good as Dall-E for prompt adherence and off training creativity.
Dall-e has great prompt understanding, you're right there my friend! 😊💫
Agreed 🤖👋
Thanks for the guide. I didn't know some of the features in recraft.
recraft is really nice if you want to make cartoons, and it's ability to do text is fantastic. however it's not 'the world's best'
Can you make a video about Hunyuan, the opensource generative video AI?
Wow some real negative commenters. Most do not understand that different people want different things and different tools.
Just because you are not a fan does not mean it is bad. There are many different jobs out there that require different tool based on different skillsets. And dropping a comment that you do not like it is rather pointless, we all have opinions, does not mean we need to all voice them.
I like to be across ALL the tools available so I can recommend the right tool for the right person, so thank you for this video and all the others. 🙏🖤
Thank you, that's a very kind and valid comment 🌟😊
from my testing recraft doesn't even come close to Flux in image generation. it has cool tools similar to Canva.
Thanks for sharing 👍
Happy to do it!
Great video :)
Glad you enjoyed it!
Def going to dig into and checkout 100!
Cool! Good to see you again 💫
nice, but still = waste of time ;)
I guess it depends on the user and their usecase 😊
true, unless weights are open to download locally, still sdxl, flux dev, maybe sd3.5 are best.
@@sebastiankamphor who is putting food in your table
🥺
Sadge? Here's a hug.
@@sebastiankamph
It's nice, but if it is not open source, no interest.
Not bad, but in your video im not really seeing anything ground breaking that makes me go 'wow flux /SD or MJ cant do anything like that' When new Generative text to image AI tools come out, why do so many inflencers test them with yet another a cute cartoon cat, or cyber punk city scape, or a pretty woman portait ? -Stuff that we know all the AI modles to date can already do just fine, SD 1.5 can do those just fine. Youre all going far to easy on them with all this safe stuff! :D Give it 'a guy throwing a ball at a lamp post' or 'a robot breaking though a wall' or 'a woman laying down or doing a high kick' or hell, just a just basic train / bus interior that doesnt look like it was designed and built by smack heads
I think people are approaching it from 2 directions currently. Researchers who want to push the limits, but have less focus on UX and UX focused tools with less interest in pushing the limits. Really hard to do both right now.
Not much inspiring about a blinding canvas. The void is described as satin black, and you create from there -- not from bright light. What year is this?
Great branding on the hat, Mr. Kamph! Thanks for all your work! Yellow is pretty great. Love.
Thank you! Much love to you too. More yellow to the people.
they are the best. I don´t understand why nobody knows them
It's not runnable locally and doesn't do NSFW, that is why.
Complete trash compare to the level of image quality offered by Ideogram V2.0 almost six months ago.
No Sebastian, please don't go in that direction. Don't let them seduce you to the dark side.
Join me son
Better than Midjourney?? Maybe if you don't know what Midjourney is or does......
"better" all depends on what you are trying to get out of the AI. They all have their strengths and weaknesses.
When I see styles to chose I'm usually out. Means you can not do your own styles and therefor useless for serious work. Also this is not testing. This is a joke.
you can create your own style by uploading 5 similar images. he says it in 07:56
you can do your own style in recraft tho
Another advertisement. Game-changer... Yeah, yeah... No, thank you. Too much to pay for, too little to get from.
cringe video, cringe channel, cringe topic.
Thank you, very kind!
cringe comment.