{"slug": "glitch-img-1-0-rgba-image-generation-editing-with-qwen-image-2-1", "title": "Glitch-Img-1.0: RGBA Image Generation & Editing with Qwen-Image-2.1", "summary": "Developer Islam Al-Nashar re-uploaded Qwen's Qwen-Image-2.1 model to Hugging Face as ISLAM-PO/Glitch-Img-1.0, keeping all weights and configs unchanged while adding documentation and a demo Space. The model supports text-to-image generation up to 2048x2048, inpainting-style editing, native RGBA transparency output, and up to 10 reference images for subject consistency, and is released under the non-commercial Qwen Research License. Al-Nashar said he uses it as the visual layer of an Arabic-first AI stack alongside his text model MasryGPT.", "body_md": "**Try it:** [huggingface.co/ISLAM-PO/Glitch-Img-1.0](https://huggingface.co/ISLAM-PO/Glitch-Img-1.0) · [image-gen demo Space](https://huggingface.co/spaces/ISLAM-PO/image-gen-demo)\n\nImage generation has moved past \"pretty pictures\" toward **controllable, editable, transparent assets**. **Glitch-Img-1.0** is my re-upload of `Qwen/Qwen-Image-2.1` under the name `ISLAM-PO/Glitch-Img-1.0`, aimed at making a strong unified text-to-image + editing model easy to discover and use from a single, well-documented repository. All weights and configs are unchanged from the original; only the README and repo IDs were updated, and the model is **Built with Qwen** under the Qwen Research License (non-commercial).\n\nThis report covers what the model does, its key capabilities, a quick-start, memory tips, and licensing notes.\n\n`Qwen/Qwen-Image-2.1`\n`text-to-image` (diffusers: `QwenImage21Pipeline`)\nFour headline improvements define the Qwen-Image-2.1 family this build inherits:\n\n| Capability | Details | \n|---|---|\n| Text-to-image | Prompts up to 2048×2048 and beyond; strong text rendering | \n| Image editing | Inpainting-style edits, background swaps, style transfer | \n| RGBA transparency | Native alpha channel generation with a documented prompt format | \n| Multi-reference | Up to 10 reference images for identity/subject consistency | \n| Local edits | Circles, painted annotations, separate masks | \n| Aspect ratios | 1:1, 4:3, 3:4, 3:2, 2:3, 16:9, 9:16 presets | \n\n**Install:**\n\n```\npip install torch>=2.4.0\npip install transformers>=5.17\npip install git+https://github.com/huggingface/diffusers\npip install accelerate pillow\n```\n\n**Text-to-image:**\n\n``` python\nimport torch\nfrom diffusers import QwenImage21Pipeline\n\npipe = QwenImage21Pipeline.from_pretrained(\n    \"ISLAM-PO/Glitch-Img-1.0\", torch_dtype=torch.bfloat16\n).to(\"cuda\")\n\nimage = pipe(\n    prompt='A neon shop sign that reads \"GLITCH IMG 1.0\", rainy night, reflections on wet pavement',\n    width=2048, height=2048,\n    num_inference_steps=40,\n    generator=torch.Generator(\"cuda\").manual_seed(42),\n).images[0]\n\nimage.save(\"t2i_example.png\")\n```\n\n**Image editing:**\n\n``` python\nimport torch\nfrom PIL import Image\nfrom diffusers import QwenImage21Pipeline\n\npipe = QwenImage21Pipeline.from_pretrained(\n    \"ISLAM-PO/Glitch-Img-1.0\", torch_dtype=torch.bfloat16\n).to(\"cuda\")\n\ninput_image = Image.open(\"input.png\")\nimage = pipe(\n    prompt=\"Change the background to a sunset beach\",\n    image=input_image,\n    num_inference_steps=40,\n    generator=torch.Generator(\"cuda\").manual_seed(42),\n).images[0]\nimage.save(\"edit_example.png\")\n```\n\n**Transparent (RGBA) generation** — use the recommended prompt format:\n\n```\nimage = pipe(\n    prompt=\"This is an RGBA image with transparency. A cute cartoon dragon sticker. \"\n           \"The image has alpha channel and the background is transparent.\",\n    width=2048, height=2048,\n    num_inference_steps=40,\n    generator=torch.Generator(\"cuda\").manual_seed(42),\n).images[0]\nimage.save(\"transparent_example.png\")\n```\n\n**Aspect ratios:**\n\n```\naspect_ratios = {\n    \"1:1\":  (2048, 2048), \"4:3\":  (2400, 1792), \"3:4\":  (1792, 2400),\n    \"3:2\":  (2528, 1696), \"2:3\":  (1696, 2528), \"16:9\": (2752, 1536),\n    \"9:16\": (1536, 2752),\n}\npipe = QwenImage21Pipeline.from_pretrained(\n    \"ISLAM-PO/Glitch-Img-1.0\", torch_dtype=torch.bfloat16\n)\npipe.enable_model_cpu_offload()\n```\n\nThis offloads model parts to CPU between steps, letting you run on smaller GPUs at the cost of speed.\n\nI use Glitch-Img as the **visual layer** behind my portfolio and product work:\n\nIt complements **MasryGPT** (text) as the two halves of an Arabic-first AI product stack: one generates the words, the other the visuals.\n\n`model-business@notice.qwencloud.com`.` NOTICE` and `LICENSE` in the repo for required attribution.\n*Author: Islam Al-Nashar — Full-Stack Developer & AI builder. [GitHub](https://github.com/aslamalkarywk7) · [Hugging Face](https://huggingface.co/ISLAM-PO) · [Portfolio](https://aslamalkarywk7.github.io/aslamalkarywk7/)*", "url": "https://wpnews.pro/news/glitch-img-1-0-rgba-image-generation-editing-with-qwen-image-2-1", "canonical_source": "https://dev.to/islam_alnashera_19f64a4e/glitch-img-10-rgba-image-generation-editing-with-qwen-image-21-1075", "published_at": "2026-10-31 13:39:27+00:00", "updated_at": "2026-10-31 13:49:55.557626+00:00", "lang": "en", "topics": ["generative-ai", "ai-tools", "artificial-intelligence"], "entities": ["Islam Al-Nashar", "Glitch-Img-1.0", "Qwen", "Qwen-Image-2.1", "Hugging Face", "MasryGPT", "diffusers"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/glitch-img-1-0-rgba-image-generation-editing-with-qwen-image-2-1", "markdown": "https://wpnews.pro/news/glitch-img-1-0-rgba-image-generation-editing-with-qwen-image-2-1.md", "text": "https://wpnews.pro/news/glitch-img-1-0-rgba-image-generation-editing-with-qwen-image-2-1.txt", "jsonld": "https://wpnews.pro/news/glitch-img-1-0-rgba-image-generation-editing-with-qwen-image-2-1.jsonld"}}