💡 DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editing

DeepGen 1.0 is a lightweight unified multimodal model with only 5B parameters (3B VLM + 2B DiT). It integrates five core capabilities—general image generation, general image editing, reasoning image generation, reasoning image editing, and text rendering—within a single model. Across multiple authoritative benchmarks, DeepGen 1.0 is competitive with competitive with or surpassing the state-of-the-art unified multimodal models that are 3× to 16× larger, achieving comprehensive performance, demonstrating that massive scaling is not the sole path to high-performance multimodal generation.

🧠 Method

Our core observation is that a lightweight model, when empowered by synergistic architecture design and data-centric training strategies, can achieve comprehensive capabilities competitive with or even surpassing much larger counterparts. To overcome the limitations of lightweight models in semantic understanding and fine-grained control, we introduce Stacked Channel Bridging (SCB), a deep alignment framework that extracts hierarchical features from multiple VLM layers and fuses them with learnable ``think tokens'' to provide the generative backbone with structured, reasoning-rich guidance. We further design a data-centric training strategy spanning three progressive stages: (1) Alignment Pre-training on large-scale image-text pairs and editing triplets to synchronize VLM and DiT representations, (2) Joint Supervised Fine-tuning on a high-quality mixture of generation, editing, and reasoning tasks to foster omni-capabilities, and (3) Reinforcement Learning with MR-GRPO, which leverages a mixture of reward functions and supervision signals, resulting in substantial gains in generation quality and alignment with human preferences, while maintaining stable training progress and avoiding visual artifacts.

📊 Benchmarks

1. General Image Generation

2. General Image Editing

3. Reasoning Image Generation

4. Reasoning Image Editing

🎨 Quantitative results

🛠️ Usage

Merge ZIP Files

To use the DeepGen checkpoints, please merge the sharded model files first. We release Pre-traning, Supervised Fine-Tuning and Reinforcement Learning checkpoints.

# Merge zip
cat DeepGen_CKPT.zip.part-* > DeepGen_CKPT.zip
# Unzip DeepGen checkpoints 
unzip DeepGen_CKPT.zip

checkpoints/
├── DeepGen_CKPT
    ├──Pretrain├──iter_200000.pth
    ├── SFT├──iter_400000.pth
    ├──RL├──MR-GDPO_final.pt 

if you want only final model state please use model.pt directly , it is same as MR-GDPO_final.pt

the Pretrain├──iter_200000.pth and SFT├──iter_400000.pth can be loaded for continuous training

⭐ Citation

@article{wang2026deepgen,
  title={DeepGen 1.0: A Lightweight Unified Multimodal Model for Advancing Image Generation and Editing},
  author={Wang, Dianyi and Li, Ruihang and Han, Feng and Ma, Chaofan and Song, Wei and Wang, Siyuan and Wang, Yibin and Xin, Yi and Liu, Hongjian and Zhang, Zhixiong and others},
  journal={arXiv preprint arXiv:2602.12205},
  year={2026}
}