Unified Multimodal Image Generation & Editing
Most tools make you juggle a separate model for each job — one to generate, another to edit, another for reasoning-heavy prompts. DeepGen 1.0 folds all of that into a single lightweight model, like a Swiss Army knife instead of a drawer full of single-use gadgets. At just 5B parameters it stays small enough to run without a data-center GPU, yet holds its own against models 3× to 16× its size. Built by the DeepGen team.
🏆 GenEval 0.87 and DPGBench 87.90 (silver among open-source models), plus #1 open-source on the WISE reasoning benchmark (0.73) — at just 5B parameters
What it does
DeepGen 1.0 is a unified multimodal image generation and editing model built on a hybrid VLM + DiT (Diffusion Transformer) architecture. With only 5B parameters (3B VLM + 2B DiT), it packs five core capabilities — general image generation, general image editing, reasoning-based generation, reasoning-based editing, and text rendering — into one model, while staying competitive with or beating models 3× to 16× larger. It is designed to show that massive scaling is not the only path to high-performance multimodal generation.
Problem it solves
- Lightweight efficiency – Achieves state-of-the-art performance at just 5B parameters, far smaller than comparable unified models (e.g., BAGEL at 14B, Hunyuan-Image-3.0 at 80B)
- Unified capabilities – Combines generation, editing, reasoning generation, reasoning editing, and text rendering in one model, eliminating the need for separate specialized models
- Semantic understanding – Overcomes limitations of lightweight models via Stacked Channel Bridging (SCB), enabling fine-grained semantic control
- Reasoning-aware generation – Supports complex, knowledge-intensive prompts that require multi-step reasoning before generating or editing an image
- Creative workflows – Suitable for design, content creation, and multimodal applications requiring both generation and editing
Input/Output
- Input:
- Text-to-image: a natural-language prompt describing the desired image
- Image-to-image: one or more reference images + a text instruction for editing or transformation (supports general editing, style transfer, reasoning-based editing, and identity preservation)
- Output: Generated or edited image
- Quality: High-fidelity, semantically accurate, fine-grained detail
- Options:
- Diffusion steps — 10–50; controls how many refinement iterations the model uses (default: 50)
- Output image height & width — default 512×512
- Random seed (Advanced) — set for reproducible results
Accuracy & Speed
General Image Generation:
Benchmark | Score | Rank |
Geneval ↑ | 0.87 | 🥈 among open-source models |
DPGBench ↑ | 87.90 | 🥈 among open-source models |
UniGenBench ↑ | 75.74 | 🥈 among open-source models |
General Image Editing:
Benchmark | Score | Rank |
GEdit-EN ↑ | 7.17 | 🥉 among open-source models |
ImgEdit ↑ | 4.14 | 🥉 among open-source models |
Reasoning Image Generation:
Benchmark | Score | Rank |
WISE ↑ | 0.73 | 🥇 among open-source models |
T2I-CoREBench ↑ | 46.5 | 🥈 among open-source models |
Reasoning Image Editing:
Benchmark | Score | Rank |
RISE ↑ | 10.8 | 🥉 among open-source models |
UniREditBench ↑ | 75.7 | 🥈 among open-source models |
Model Source
- Hugging Face
Compliance & Provenance
Provider | Open-source |
Provider type | Specialized |
License | |
EU AI Act risk class | Limited Risk |
Art. 50 transparency | Required — outputs are marked. See AI Policy §2. |
Region availability | Available globally |
Training data summary | Pending — provider has not yet published per Art. 53(d) |
For more on how we classify models and mark outputs, see our AI Policy.