OpenAI’s DALL-E 3-Like AI For Free, Forever!

❤️ Check out Weights & Biases and sign up for a free demo here:

Flux is available here:

Try it:
1.
2.

Run it yourself at home:
1.
2.

Image credit:

📝 My paper on simulations that look almost like reality is available for free here:

Or this is the orig. Nature Physics link with clickable citations:

🙏 We would like to thank our generous Patreon supporters who make Two Minute Papers possible:
Alex Balfanz, Alex Haro, B Shang, Benji Rabhan, Gaston Ingaramo, Gordon Child, John Le, Kyle Davis, Loyal Alchemist, Lukas Biewald, Martin, Michael Albrecht, Michael Tedder, Owen Skarpness, Richard Sundvall, Taras Bobrovytsky, Ted Johnson, Thomas Krcmar, Tybie Fitzhugh, Ueli Gallizzi.
If you wish to appear here or pick up other perks, click here:

My research:
X/Twitter:
Thumbnail design: Felícia Zsolnai-Fehér –

Joe Lilli
 

  • @devonbain335 says:

    Loving the breakdowns as always.

  • @MrYTThor says:

    What a time to be alive!

  • @ConnorisseurYT says:

    1 minute in and I already dropped my papers

  • @whatworld6834 says:

    I ahve to admit, I was skeptical about flux, but I got it to run on my local machine. It maxes out my system but on first load it takes a bout 416 secs to render an image, but after that just 100 secs. Though I am using the slimed down model Schenlle but still great results! Better than what I got with SD

  • @froilen13 says:

    Waiting for Loras and controlnet

  • @jonmichaelgalindo says:

    This is much better than DallE3 in my physics tests. It passed “Hand holding a pair of scissors casting a shadow on a wall” 😊🎉

  • @EVILBUNNY28 says:

    I love this, hopefully it makes Dale-3 and Midjourney creators rethink how much they charge for their own models.

  • @AlmogKarif says:

    free? why is it say then “Your request will cost $0.05 per megapixel. For $1 you can run this model approximately 20 times.”
    ???

    • @ImNotQualifiedToSayThisBut says:

      Because you can run it locally?

    • @bars7897 says:

      I’ve used it once before seeing that writing. Should I be concerned?

    • @ImNotQualifiedToSayThisBut says:

      @@bars7897 about what?

    • @UMessedUp says:

      @@bars7897 no lol its just for when you run out of generations and you need to pay for more. the first 40 are free

    • @IceMetalPunk says:

      That’s only when using the hosted UI. There’s also one hosted on HuggingFace that’s totally free, plus it’s open-source and open-weights, so you can run it entirely free locally if you have a decent enough graphics card. (Seems like the recommendation is at least 12GB VRAM, preferably more; someone got it to run on 6GB but it overheated their card in the process….)

  • @Dave-rd6sp says:

    Unfortunately, the Dev model, the fine-tunable one, is non-commercial. But it looks like the Schnell model (the Apache 2.0 one) has been un-distilled and made fine-tunable.

  • @vilchico says:

    Hello, do you have the project files to follow along?

  • @vi6ddarkking says:

    A group of AI researchers leave Stability recruit a friend beat Stability at their own game.
    Love to see it.

  • @hinro says:

    technically you can run it at home but currently it is sitting at 24gb which puts it outside of most consumer cards. hopefully soon though the optimizations come in and it can really be run at home.

  • @williamlocke6811 says:

    I can’t stop playing with this, it’s literally amazing! Imagine the fine-tunes! :O

  • @shaderbytes says:

    it failed dramatically when I tested it yesterday , no matter writting many key words like , CG , Cartoon , styled like Pixar , Rendered .. it just spat out photo realistic. I even tried different algorithms. On top of it not interpretting my text for style at all, it also drew almost all the images it did make with one of the hands having 6 or 7 fingers each time..

  • @RealityRogue says:

    I end up using “Niji 6” of midjourney for it’s amazing creative quality when coming up with stylized characters. I don’t know if Flux would come close, as this seems mostly similar to midjourney’s V6.1

  • @bars7897 says:

    One of the links asks for 0.05 cents per megapixel. And I’ve used it before I read that. Am I forever in debt now? Should I panic? Was this a scam?

  • @nixes1636 says:

    This is amazing. Thank you so much for sharing it. I wouldn’t have found it otherwise.

  • @tbzor says:

    The text plus pictures is so clean, could one specify font though?

  • @DonCarnage42 says:

    Looking good. Did an apple pie with letters on top using Schnell, and it worked perfectly the second time.

  • @emanuelescarsella3124 says:

    I was able to try it out before everyone made it impossible to try😅
    And with just 4 steps the results are already great, especially when it comes to prompt comprehension it’s better than everything else at the moment… But I wasn’t able to try it with more steps, I’ll surely get back to it

  • >