Z-Image-Re-Turbo is a text-to-image generation model that has been optimized for de-reduction and re-acceleration based on the Z-Image-De-Turbo model. This model aims to balance the convenience during training and the speed during inference. It restores the fast generation ability close to the original Turbo model while maintaining the same training-friendly features as Z-Image-De-Turbo, enabling it to be perfectly compatible with a large number of trained LoRA models in the Z-Image ecosystem.

MultimodalDiffusers

INTELLECT-3 is a Mixture of Experts (MoE) inference model developed by Prime Intellect with a parameter scale of 106B (12B active parameters). It is supervised and fine-tuned based on GLM-4.5-Air-Base and trained through large-scale reinforcement learning, showing excellent performance in mathematics, coding, and inference tasks.

Natural Language ProcessingGguf

This is the GGUF quantization format version of the NousResearch/Hermes-4.3-36B model, quantized by MaziyarPanahi. GGUF is an efficient model format designed to optimize local deployment and inference performance, supporting multiple quantization levels (2-bit to 8-bit), and suitable for a wide range of text generation tasks.

Natural Language ProcessingGguf

GigaChat3-10B-A1.8B is an efficient dialogue model in the GigaChat series. Based on the Mixture-of-Experts (MoE) architecture, it has a total of 10 billion parameters and 1.8 billion active parameters. It adopts the innovative Multi-head Latent Attention (MLA) and Multi-token Prediction (MTP) technologies, aiming to optimize inference throughput and generation speed. The model is trained on 20T tokens of diverse data and supports 10 languages including Chinese, suitable for dialogue scenarios requiring quick responses.

Natural Language ProcessingGgufMultiple Languages

This is the GGUF quantized version of the large language model Mistral-Large-3-675B-Instruct-2512 developed by Mistral AI. The original model has 675 billion parameters and is designed for instruction-following tasks. This project uses the llama.cpp tool and combines it with the imatrix calibration dataset to generate more than 20 different precision quantized model files from Q8_0 to IQ1_S, aiming to balance model performance, inference speed, and storage/memory usage, enabling it to run on a wider range of hardware.

Natural Language ProcessingGgufMultiple Languages

This is a lightweight text generation model based on the Meta LLaMA 3 8B model. It is efficiently fine-tuned using the Unsloth framework and converted to the GGUF format. The model is optimized for the Alpaca dataset, suitable for local deployment and inference, especially in resource-constrained environments.

Natural Language ProcessingGguf

This project provides a text generation model fine-tuned based on the Microsoft Phi-3.5-mini-instruct model, which has been converted to the GGUF format and is suitable for the llama.cpp inference framework. The model was fine-tuned on the philschmid/guanaco-sharegpt-style dataset, optimizing its instruction-following and dialogue capabilities, and is suitable for efficient text generation tasks in resource-constrained environments.

Natural Language ProcessingGguf

Z-Image is an efficient basic image generation model with 6 billion parameters, specifically designed to address the efficiency and quality issues in the field of image generation. Its distilled version, Z-Image-Turbo, can reach or exceed the leading competitors with only 8 function evaluations. It can achieve sub-second inference latency on enterprise-level H800 GPUs and run on consumer-grade devices with 16G VRAM.

Computer VisionDiffusersMultiple Languages

Z-Image-Turbo is an image processing model quantized in FP8 E5M2 and E4M3FN formats. It is optimized based on the original Tongyi-MAI/Z-Image-Turbo model, significantly reducing the model size and inference resource requirements while maintaining performance.

MultimodalDiffusersEnglish

An optimized image generation model based on black-forest-labs/FLUX.2-dev, supporting image-to-image generation and editing tasks. It uses fp8_scaled quantization technology to improve inference speed and is particularly suitable for use on low-VRAM GPUs.

This model is an MLX format conversion version of Kimi-Linear-48B-A3B-Instruct, optimized for Apple Silicon devices such as the Apple Mac Studio. It is a large language model with 48 billion parameters, supporting instruction following and suitable for local inference and conversation tasks.

Natural Language ProcessingMlx

This is a quantized version of the squ11z1's Hypnos-i1-8B model, which is quantized using the llama.cpp tool and a specific dataset. The model is based on the Llama-3 architecture, offering multiple quantization type options and supporting text generation tasks. It is suitable for various application scenarios such as inference and dialogue.

Natural Language ProcessingGgufEnglish

This repository contains a 4-step distilled model optimized for HunyuanVideo-1.5, which can achieve ultra-fast 4-step inference without using CFG (Classifier-Free Guidance), significantly reducing the generation time while maintaining high-quality video output.

MultimodalDiffusers

This is an MXFP4_MOE imatrix quantized version based on the Huihui-MiroThinker-v1.0-30B model, specifically optimized for text generation tasks. It significantly reduces the model size and inference cost while maintaining model performance.

Natural Language ProcessingGguf

This is an optimized version based on the original DeepSeek-OCR model, specifically an OCR model that supports inference on Apple Metal Performance Shaders (MPS) and CPU. It can extract text from images and convert it into a structured format, supporting multilingual document recognition.

MultimodalTransformersOther

GLM-4.5-Iceblink-v2-106B-A12B-FP8 is a version based on the GLM-4.5-Iceblink-v2-106B-A12B model, using the most advanced mixture of experts quantization method for FP8 quantization. This model is specifically optimized for Ada, Hopper, or Blackwell series GPUs that support hardware FP8, significantly improving inference efficiency while maintaining high-quality output.

Natural Language ProcessingSafetensors

This is a quantized version based on the Huihui-MiroThinker-v1.0-8B model, specifically optimized for text generation tasks. It improves inference efficiency through quantization technology while maintaining the model's performance.

Natural Language ProcessingGguf

This is a version of the MiroThinker-v1.0-30B model quantized using MXFP4_MOE imatrix, implemented based on mradermacher's imatrix. This quantized model maintains the text generation ability of the original model while improving inference efficiency through quantization technology. It is suitable for various application scenarios that require efficient text generation.

Natural Language ProcessingGguf

GigaChat3-10B-A1.8B-base is the basic pre-trained model of the GigaChat series, adopting the Mixture of Experts (MoE) architecture with a total of 10 billion parameters and 1.8 billion active parameters. The model integrates Multi-Head Latent Attention (MLA) and Multi-Token Prediction (MTP) technologies, and has the advantage of high throughput during inference.

Natural Language ProcessingSafetensorsMultiple Languages

This project provides a quantized version of the cerebras/MiniMax-M2-REAP-172B-A10B model, aiming to make knowledge accessible to the public. This is a large language model with 172 billion parameters, which has been optimized and quantized to reduce deployment costs and improve inference efficiency.

Natural Language ProcessingGguf