cs.CVNov 5, 2026

WeLM-Image: A Unified Image Generation Model towards Visual Intelligence

Authors: Canmiao Fu, Chen Li, Guohui Zhang, Hao Lu, Hongyang Wei, Huadong Wu, Huaisong Zhang, Hubery Yin, 

Organizations: WeChat Vision, Tencent Inc.

Abstract

We present WeLM-Image, a unified image foundation model that integrates multimodal understanding into text-to-image generation, instruction-guided editing, and multi-reference synthesis. Designed for single-GPU deployment in WeChat's large-scale consumer applications, the model processes text, reference images, and target latents as one interleaved sequence. A Mixture-of-Transformers (MoT) backbone combines shared attention with independent understanding and generation parameters to limit optimization interference. Equipping each path with Mixture-of-Experts (MoE) modules provides sparse capacity, with 31.7B stored parameters and approximately 5.4B generation parameters activated per target latent token. A generation-oriented data engine combines a knowledge-grounded visual vocabulary with large-scale visual clustering to balance semantic coverage and editing tasks, while training-time evaluation guides iterative data refinement. Bilingual annotations and verified editing supervision support progressive training up to native 1.5K resolution. We optimize generation and editing separately with rubric- and checklist-based rewards, then consolidate the specialized policies through noise-aware on-policy distillation. In the reported overall comparisons, the 1536-resolution model ranks first among the evaluated open-source models on PRISM-Bench, Qwen-Image-Bench, and UnicBench. MeanFlow-based few-step distillation yields WeLM-Image-Turbo, which generates a 1024×10241024\times1024 image in 4.98,s on a single NVIDIA H20 GPU using eight guidance-free steps.

Figures & tables

Explore similar work

CardsList
  1. Qwen-Image-2.0 Technical Report

    May 11, 2026Bing Zhao, Chenfei Wu, Deqing Li +72T2I GenerationCompositional T2I Generation

  2. Ask, Solve, Generate: Self-Evolving Unified Multimodal Understanding and Generation via Self-Consistency Rewards

    Jun 25, 2026Ritesh Thawkar, Shravan Venkatraman, Omkar Thawakar +5Unified Multimodal ModelsMultimodal Generation