• God@sh.itjust.worksM
    link
    fedilink
    English
    arrow-up
    13
    ·
    1 year ago

    i pay for MJ because i am very rich 😎 (actually i just have bad self control and will enter my card on any site that promises me a tiny droplet of happiness)

  • TheGruesomeTwosome@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    12
    ·
    1 year ago

    Yes! Bing Image Creator is totally free and used DALL-E. It’s a one and done kinda deal, you’re not tweaking and getting as specific as some others, but you will get some amazing results with shorter prompts like “draw a realistic hamster putting out a fire with a hose, with a city background”

    • nieceandtows@lemmy.worldOP
      link
      fedilink
      English
      arrow-up
      4
      ·
      1 year ago

      Thanks, it worked. I see a 100 counting down with each image I create. Will it get replenished automatically?

      • TheGruesomeTwosome@sh.itjust.works
        link
        fedilink
        English
        arrow-up
        2
        ·
        1 year ago

        Ah, so Bing will only let you go so far with the AI, which is 30 messages, up from 20. This is really just to stop things going too far I believe, as the tech really is nuts. It’ll also often catch itself getting too lewd or whatever and cut itself off. Bing AI is way more “out there” than some others

    • Jord@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      3
      ·
      1 year ago

      Thanks for the suggestion - I just tried this and made a pretty cool image for my latest Pathfinder character!

    • planish@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 year ago

      Any news of like a llama.cpp equivalent for SD? It would be handy to be able to slowly run it without a GPU, and maybe competitive with other free options in terms of images generated per day.

        • planish@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          3
          ·
          1 year ago

          It’s a portable, standard-C++, CPU-based implementation of the code to do inference (i.e. text generation) with the LLaMA language model. You get a command line command that takes text and the model and eventually outputs more text.

          You could do the same thing and run Stable Diffusion off of the CPU at some relatively slow speed, but I don’t know if anyone has code for it.

            • planish@sh.itjust.works
              link
              fedilink
              English
              arrow-up
              2
              ·
              1 year ago

              Wow, that claims to be really fast on CPU actually. Why aren’t people using this all the time instead of the annoying services?

              • Inductor@feddit.de
                link
                fedilink
                English
                arrow-up
                1
                ·
                edit-2
                1 year ago

                That guide refrences this version for more CPU performance. I haven’t tried using CPU, but from my experience with raytraced rendering on a CPU, it’s probably very slow compared to GPU. It might be faster than online services with a queue though.

    • db0@lemmy.dbzer0.com
      link
      fedilink
      English
      arrow-up
      6
      ·
      1 year ago

      Came here to say that. We also have multiplex other UIs so you can find the one that works best for you. 4 minutes might also be pushing it, even for anonymous. You can make an account however and go much faster. And you can easily get a lot more kudos by contributing to the community

    • planish@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      3
      ·
      1 year ago

      Another vote for the AI Horde people. They seem to be a legit volunteer operation not about to turn startup, and the kudos economy they are running is pretty neat and well thought out.

    • azayrahmad@sh.itjust.works
      link
      fedilink
      English
      arrow-up
      3
      ·
      edit-2
      1 year ago

      I myself use Patience.ai. Although I’m not sure what kudos do.

      The queues are mostly on popular models (anime, photorealistic, the like), so perhaps you want to use more niche models (Supermarionation? Can’t believe it exists). Also instead of showing time estimation, Patience.ai would show your queue number instead (322nd image in queue).

    • ouigol@lemmy.world
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 year ago

      That’s dalle mini, right? Last time I used it it only produced nightmarish images lol. Maybe it’s better now?

        • azayrahmad@sh.itjust.works
          link
          fedilink
          English
          arrow-up
          3
          ·
          1 year ago

          Honestly the strength of Craiyon is not in aesthetics, but in accuracy. Craiyon is trained using unfiltered images from internet, unlike other AI image generators that use curated images.

          In my experience, StableDiffusion or even Dall-E often misinterpret my prompt, Craiyon rarely disappoints me. Dall-E has prettier result though.

          I usually generate images using Craiyon first, then I use the result as starting image for mage.space/nightcafe/patience.ai.

  • azayrahmad@sh.itjust.works
    link
    fedilink
    English
    arrow-up
    5
    ·
    edit-2
    1 year ago
    • NightCafe is not exactly free, but you get to claim 5 credits everyday. Generating images cost 1 credit or more depending on model and steps. You can accumulate credits. It’s much better than 15 credit monthly of DALL-E.

    • Mage.Space is free for StableDiffusion models, but offers paid upgrade to access other advanced models and NSFW art.

    • B Discover is a mobile app that could generate AI images for free as well.

    • Craiyon has been mentioned here, it is also free, although the generation time can take quite long (they offer paid upgrade for this), and the results are often not pretty. Each resulting image is also watermarked with Craiyon logo in the corner. However I must admit that Craiyon model is very accurate, whatever weird prompt you give it will try to obey, instead of just giving generic pretty images.

    • nieceandtows@lemmy.worldOP
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 year ago

      I’m getting an error trying to submit a prompt. " Failed to generate images due to: SyntaxError: JSON.parse: unexpected character at line 1 column 1 of the JSON data " Any idea?

      • bdiddy@lemmy.one
        link
        fedilink
        English
        arrow-up
        2
        arrow-down
        31
        ·
        1 year ago

        It’s finiky sometimes I just gotta come back later and it works fine