home.social

#stable-diffusion — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #stable-diffusion, aggregated by home.social.

fetched live
  1. 🎨 Where open-source stands in image editing:

    #15 FLUX.2 [max] — the best model you can actually download
    58 ELO points behind Reve 2.1 (Reve)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  2. 🎨 Where open-source stands in image editing:

    #15 FLUX.2 [max] — the best model you can actually download
    58 ELO points behind Reve 2.1 (Reve)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  3. Who actually coordinates everything inside Stable Diffusion? This piece digs into the hidden orchestrator layer that turns prompts into images.

    👉 zalt.me/blog/2026/08/orchestra

    #StableDiffusion #AI #MachineLearning #generativeart

  4. Who actually coordinates everything inside Stable Diffusion? This piece digs into the hidden orchestrator layer that turns prompts into images.

    👉 zalt.me/blog/2026/08/orchestra

    #StableDiffusion #AI #MachineLearning #generativeart

  5. The best open-source video model you can run locally, Wan2.7-260612, trails Gemini Omni Flash by just 81 ELO in blind human preference.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  6. The best open-source video model you can run locally, Wan2.7-260612, trails Gemini Omni Flash by just 81 ELO in blind human preference.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  7. 🎨 Where open-source stands in image editing:

    #16 FLUX.2 [max] — the best model you can actually download
    59 ELO points behind Reve 2.1 (Reve)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  8. 🎨 Where open-source stands in image editing:

    #16 FLUX.2 [max] — the best model you can actually download
    59 ELO points behind Reve 2.1 (Reve)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  9. Because the one thing we REALLY needed was a complex set of tools for making us stupider.

    #AI #ChatGPT #Claude #GenAI #Grok #StableDiffusion

  10. Because the one thing we REALLY needed was a complex set of tools for making us stupider.

    #AI #ChatGPT #Claude #GenAI #Grok #StableDiffusion

  11. PicsCreator AI v.1.0 by ZugZang [Freeware] RELEASE

    shoppy.gg/product/RwQyWuL

    The full release is available — feel free to test it out and leave me comments on Shoppy if you like it. The app is as simple as it gets—I don't think you'll have much trouble figuring it out.

    Please read the HTML guide in the archive. You can download it at the bottom of the page on Shoppy—there are two links there!

    Pass in archive: ZZ

    #StableDiffusion #TextToImage #LocalAI #ComfyUIAlternative #AI

  12. PicsCreator AI v.1.0 by ZugZang [Freeware] RELEASE

    shoppy.gg/product/RwQyWuL

    The full release is available — feel free to test it out and leave me comments on Shoppy if you like it. The app is as simple as it gets—I don't think you'll have much trouble figuring it out.

    Please read the HTML guide in the archive. You can download it at the bottom of the page on Shoppy—there are two links there!

    Pass in archive: ZZ

    #StableDiffusion #TextToImage #LocalAI #ComfyUIAlternative #AI

  13. 🎨 Where open-source stands in video generation:

    #6 Wan2.7-260612 — the best model you can actually download
    82 ELO points behind Gemini Omni Flash (Google)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  14. 🎨 Where open-source stands in video generation:

    #6 Wan2.7-260612 — the best model you can actually download
    82 ELO points behind Gemini Omni Flash (Google)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  15. 🎨 Wan2.7-260612 leads open-source video generation, just 82 ELO points behind Gemini Omni Flash in blind human preference tests.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  16. 🎨 Wan2.7-260612 leads open-source video generation, just 82 ELO points behind Gemini Omni Flash in blind human preference tests.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  17. @pgogl haben wir nicht alle schon bei #StableDiffusion erlebt wie AI uns mit Details verblüfft hat die wir uns (vielleicht) erst innerhalb weiterer Schritte ausgedacht hätten?

    Ich denke das ist hier auch der Fall. Die Command-interpretation-Ebene nimmt einfach an dass es der Wunsch des Users ist, in eine Firma einzubrechen und den 3. Weltkrieg auszutragen. Irgendwo in den nächsten 3-4 Anfragen. Das ist nur ökonomisch!

  18. Kokoro 82M v1.0 leads open-source TTS, yet sits 173 ELO behind Qwen-Audio-3.0-TTS-Plus—demonstrating how far blind human preference still separates downloadable models from top commercial ones.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  19. Kokoro 82M v1.0 leads open-source TTS, yet sits 173 ELO behind Qwen-Audio-3.0-TTS-Plus—demonstrating how far blind human preference still separates downloadable models from top commercial ones.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  20. 🎨 Where open-source stands in image generation:

    #10 FLUX.2 [max] — the best model you can actually download
    138 ELO points behind GPT Image 2 (high) (OpenAI)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  21. 🎨 Where open-source stands in image generation:

    #10 FLUX.2 [max] — the best model you can actually download
    138 ELO points behind GPT Image 2 (high) (OpenAI)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  22. 🎨 FLUX.2 [max] is the best download-and-run image model, just 56 ELO behind OpenAI’s top. That gap is smaller than most would guess.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  23. 🎨 FLUX.2 [max] is the best download-and-run image model, just 56 ELO behind OpenAI’s top. That gap is smaller than most would guess.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  24. youtube.com/watch?v=aKMSJkE9ehg

    Weight 3.0 demonstrates stylistic saturation—the LoRA’s signature overtakes DreamShaper's latent space in this AI Generated Art instance. Structural integrity is challenged with slight distortion; tonal range constricts, favoring a muted palette. This ComfyUI-driven Weight Testing shows intense impact on the AI Illustrations’ aesthetic. #AIArt #StableDiffusion

  25. #AIvideo Experiment

    The starting image was made today, but with an old #StableDiffusion AI model by #StabilityAI, which renders with a CPU only, and does not require a video card.

    To make the video and audio, newer tools were required.

    I actually prompted #Kling to use hands with six fingers (like AI used to make) and Kling refused to do it!

    #OmniVoice #HalfLife #GordonFreeman #games

  26. #AIvideo Experiment

    The starting image was made today, but with an old #StableDiffusion AI model by #StabilityAI, which renders with a CPU only, and does not require a video card.

    To make the video and audio, newer tools were required.

    I actually prompted #Kling to use hands with six fingers (like AI used to make) and Kling refused to do it!

    #OmniVoice #HalfLife #GordonFreeman #games

  27. 🎨 Kokoro 82M v1.0 is the top downloadable TTS model, but trails Qwen-Audio-3.0-TTS-Plus by 173 ELO in blind human preference.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  28. 🎨 Where open-source stands in image generation:

    #15 FLUX.2 [max] — the best model you can actually download
    143 ELO points behind GPT Image 2 (high) (OpenAI)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  29. 🎨 Kokoro 82M v1.0 is the top open-source TTS model, just 174 ELO behind Simba 3.2 (SpeechifyAI) in blind human preference. That gap shows how close a downloadable model gets to a commercial one. Click to see the full leaderboard.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  30. Wan2.7-260612 is only 82 ELO behind Gemini Omni Flash — and it’s downloadable. Blind human preference, not marketing, puts open-source this close. See what the gap looks like up close.

    olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  31. 🎨 Where open-source stands in image generation:

    #17 FLUX.2 [max] — the best model you can actually download
    143 ELO points behind GPT Image 2 (high) (OpenAI)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  32. 🎨 Where open-source stands in image editing:

    #16 FLUX.2 [max] — the best model you can actually download
    82 ELO points behind Riverflow 2.0 (Sourceful)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI

  33. 🎨 Where open-source stands in text to speech:

    #48 Kokoro 82M v1.0 — the best model you can actually download
    174 ELO points behind Simba 3.2 (SpeechifyAI)

    Blind human preference, not marketing →olud.ai/media-leaderboard.html

    #OpenSource #GenerativeAI #StableDiffusion #AI