Quantization here changes how the transformer is packaged, not the underlying editing-and-generation workflow: one merged adapter serves both tasks, while the text encoder and VAE remain the untouched Qwen-Image-2.1 components. That separation makes it possible to choose a smaller or faster-to-load build without replacing the rest of the pipeline.
AI Model2026
Qwen-Image-2.1-Uncensored-Omni-Quantized
Merges a rank-128 Omni LoRA into Qwen-Image-2.1 for text-to-image generation and natural-language image editing, with quantized builds for different runtimes and hardware. The base model’s text encoder, processor, and VAE are required separately.
Introduction
Information
- Websitehuggingface.co
- OrganizationsChatR Labs
- Published date2026/11/03
Categories
More Items
AI Model2026
Handles long-horizon agent tasks across codebase development, project builds, and visual work. It combines native visual tokens with large-scale agentic reinforcement learning and supports a one-million-token context.
AI Model2026
Runs a 145B agent model on a single consumer GPU using sparse latent MoE. The NVFP4 build reports 98.3 tok/s agent decoding on an RTX 5090 D, while also relying on host RAM.
AI Model2026
A general-purpose multimodal model that reads text and images and generates text. Its sparse MoE has 594B total parameters and activates 17B per token, with up to a 1M-token context for agentic, coding, and visual tasks; local inference requires a large multi-GPU cluster.