5 ms·

You can get much better results with Ideogram 2 (also free): https://ideogram.ai/assets/image/lossless/response/vF81gKjHSA6fYUSxyzzljQ https://ideogram.ai/asse

• renewiltord 2y ago
  Great tip. The text handling here is far superior.
  • layer8 2y ago
    The “1”s are still inconsistent, and of course the numbers are all wrong.
    • skybrian 2y ago
      This seems like a good idea for a contest.
• vunderba 2y ago
  You can also get reasonably close with an open model that you can run locally (flux dev). https://replicate.com/p/xm41nvz05drm00chsywb6am7f0 https://replicate.com/p/xm41nvz05drm00chsywb6am7f0 https://replicate.com/p/kdw8bnkj39rm40chsyzbyg5e04 https://replicate.com/p/kdw8bnkj39rm40chsyzbyg5e04 But of course anyone who has even a passing familiarity with scrabble is going to be able to tell that something's off.
  • GaggiX 2y ago
    The biggest problem with the default Flux model is that it generates images with that strong AI look, probably caused by the distillation of the CFG. You should try some LoRAs for this, and also prompt the model to generate the rack that holds the letters.
    • vunderba 2y ago
      Good point. I have a comfyui setup for it but its super basic right now just the diffusion model / clip loader / vae. Another thing you've probably noticed is that 99% of images from Flux tend to have that classic narrow depth of field look. I've seen people occasionally be able to get around it with pretty amusing prompt tokens like "instagram photo, selfie, gopro, etc." though.