전체 검색

AI 모델 21

• GLM-Image↗

  zai-org

  GLM-Image is an image generation model adopts a hybrid autoregressive + diffusion decoder architecture. In general image generation quality, GLM‑Image aligns with mainstream latent diffusion approaches, but it shows significant advantages in text-rendering and knowledge‑intensive generation scenarios. It performs especially well in tasks requiring precise semantic understanding and complex information expression, while maintaining strong capabilities in high‑fidelity and fine‑grained detail generation. In addition to text‑to‑image generation, GLM‑Image also supports a rich set of image‑to‑image tasks including image editing, style transfer, identity‑preserving generation, and multi‑subject consistency. + Autoregressive generator: a 9B-parameter model initialized from GLM-4-9B-0414, with an expanded vocabulary to incorporate visual tokens. The model first generates a compact encoding of approximately 256 tokens, then expands to 1K–4K tokens, corresponding to 1K–2K high-resolution image outputs. + Diffusion Decoder: a 7B-parameter decoder based on a single-stream DiT architecture for latent-space image decoding. It is equipped with a Glyph Encoder text module, significantly improving accurate text rendering within images. Post-training with decoupled reinforcement learning: the model introduces a fine-grained, modular feedback strategy using the GRPO algorithm, substantially enhancing both semantic understanding and visual detail quality.

• microsoft-AesCode-8B-GGUF↗

  prithivMLmods

  > **AesCode-8B** is Microsoft's 8B vision-language model, built on Qwen3-VL-8B-Instruct, for generating information-rich visual artifacts such as slides, posters, dashboards and reports as structured, editable HTML/CSS. It addresses a gap between two kinds of generator: code models can't see how layout, hierarchy and color come together on the canvas, while image generators compose attractive pages but misrender text, numbers and logical relationships. AesCode therefore takes an image generated from the same prompt as an aesthetic reference and follows the prompt for the required content. Tables come out as HTML table structures and charts as ECharts specifications, so both stay inspectable and verifiable. Training was cold-start SFT on 3,000 demonstrations, followed by GDPO reinforcement learning for 400 steps over 7,408 prompts, using verl and no separately trained reward model. Each target is represented as a design graph, and the reward is split into seven decoupled channels: execution, text, boundary, tablechart, layout, whitespace and design. Deterministic verifiers score what can be parsed from the code and its rendering, and a VLM judge scores the rest against a sample-specific Visual Graph Rubric. On the card's 300 infographic samples, AesCode-8B scores 82.94 Overall (Rule 90.07, Visual 75.80). That is above the reference-conditioned GPT-5.5 (81.28) and Claude Opus 4.8 (80.39), and 31.0 points above its reference-conditioned Qwen3-VL-8B backbone. The 32B companion model leads at 85.89. Style is a ceiling for every system, with no model above 60. The reference image is optional, since withholding it costs AesCode-8B only 1.00 Visual point, versus 19.55 for the base model, though quality is highest with it. The model loads through standard Transformers or vLLM, with the training code, verifier stack and rubric builder released on GitHub. The paper is still listed as under review. It is released under Apache 2.0.

• SD_PixelArt_SpriteSheet_Generator↗

  Onodofthenorth

  This Stable diffusion checkpoint allows you to generate pixel art sprite sheets from four different angles. These first images are my results after merging this model with another model trained on my wife. merging another model with this one is the easiest way to get a consistent character with each view. still requires a bit of playing around with settings in img2img to get them how you want. for left and right, I suggest picking your best result and mirroring. after you are satisfied take your photo into photoshop or Krita, remove the background, and scale to the desired size. after this you can scale back up to display your results; this also clears up some of the color murkiness in the initial outputs. This model can be used just like any other Stable Diffusion model. For more information, please have a look at the Stable Diffusion. ___ ___ For the front view use "PixelartFSS" ___ ___ For the right view use "PixelartRSS" ___ ___ For the back view use "PixelartBSS" ___ ___ For the left view use "PixelartLSS" ___ ___ These are random results from the unmerged model ___ ___ here's a result from a merge with my Hermione model ___ ___ here's a result from a merge with my cat girl model

• marigold-normals-v1-1↗

  prs-eth

  This is a model card for the `marigold-normals-v1-1` model for monocular normals estimation from a single image. The model is fine-tuned from the `stable-diffusion-2` model as described in our papers: - CVPR'2024 paper titled "Repurposing Diffusion-Based Image Generators for Monocular Depth Estimation" - Journal extension titled "Marigold: Affordable Adaptation of Diffusion-Based Image Generators for Image Analysis"

• minecraft-skin-generator-sdxl↗

  monadical-labs

  Monadical is pleased to announce the official release of the Minecraft Skin Generator XL model. We had previously released the Minecraft Skin Generator model based upon Stable Diffusion 2. This new model offers significant improvements over the last generation of models. 1. **Upgrade to Stable Diffusion XL** - Our model is now based upon the Stable Diffusion XL model, which greatly improves the quality of generated skins when compared to previous models. 1. **Transparent Layer Support** - The new model now supports the transparency layer in the hair and helmet section of the skin. * 'Saul Goodman from the TV show "Better Call Saul"'

• NSFW-gen-v2↗

  UnfilteredAI

  **Description:** NSFW-gen is a text-to-image generator developed by UnfilteredAI. This model is designed to generate all kinds of images, including explicit and NSFW (Not Safe For Work) images from textual inputs. **Features:** - **Uncensored Output:** The model produces uncensored and potentially explicit images based on textual inputs. - **Tensor Type:** Operates with FP16 tensor type for optimized performance and efficiency. - **Model Size:** With 3.47 billion parameters, the model offers a vast capacity for learning and generating diverse imagery. - **3D Style Rendering:** The model now includes 3D style/image rendering capability to generate more realistic images. (Use 3d, 3d style in your prompt) **Usage Guidelines:** - **Responsible Use:** Exercise discretion and responsibility when generating content with this model. - **Age Restriction:** Due to the explicit nature of the generated content, usage is restricted to individuals over the legal age in their jurisdiction.

• Hunyuan3D-2mini↗

  腾讯

  This repository contains the models of the paper Hunyuan3D 2.0: Scaling Diffusion Models for High Resolution Textured 3D Assets Generation. **Hunyuan3D-2mini** contains a 0.6B shape generator, which is smaller and faster than the previous 1.1B one. For code and more details on how to use it, refer to the Github repository.

• Cloudflare_clef-GGUF↗

  bartowski

  **Model details:** - Parameter count: 27B (source checkpoint) - Input support: text, image (with mmproj file) - details - Decision model: yes (clef) - details - Speculative decoding: no - imatrix: no (not possible for a decision model) - details This is a decision model, not a text generator: it reads a state (text or JSON) together with a schema of typed questions and hands back a probability for every allowed option of every question, all in a single forward pass. It never generates text, so there's no chat, no completion endpoint, and no prompt format to worry about. llama.cpp serves it through llama-server's `/v1/systemone` endpoint (the TypeSafe System One API). The full request and response shapes are in the llama-server README, but here's what a request looks like:

• MOSS-Image-Base-8-steps-dmd↗

  lodestones

  Work-in-progress distillation of OpenMOSS-Team/MOSS-Image-Base (50 steps, CFG 3) to 8 network evaluations, no CFG. Data-space DMD (arXiv 2610.12102): the generator iterates on its own clean estimate from fixed noise. Trained on compact-JSON `scene_graph` captions. Load `lora.safetensors` on the stock MOSS-Image-Base transformer (e.g. a ComfyUI LoRA loader, strength 1) and sample with the model's usual flow-matching Euler sampler, shift 5, **CFG 1**, 4-8 steps. Nothing else from this repo is needed. `transformer/` is a stock `MOSSImageTransformer` (load it with the release `pipeline_moss_image.py`); the text encoder, tokenizer and VAE are the base repo's.

• Cosmos3-Edge↗

  NVIDIA

  **Model Collection** | **Code** | **White Paper** | **Website** Cosmos3 is a collection of Omnimodal world models capable of generating dynamic, high-quality video, image, audio, and action commands from combinations of text, image, video, and action trajectory inputs. It serves as a foundational building block for a broad range of Physical AI applications and research spanning world understanding, world generation, simulation, and embodied policy learning. > **Update — August 25, 2026:** The Cosmos3-Edge generator checkpoint, runtime defaults, usage examples, and benchmark results have been updated. Users pulling from `main` should refresh their local snapshot. See the update announcement for compatibility and reproducibility details.