Lumina-DiMOO: An Omni Diffusion Large Language Model for Multi-Modal Generation and Understanding

2025/10/07 by Xin Yi, Xin, Yi, Qi Qin +58 · 19 citations
Computer Science · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Natural Language Processing Techniques #Topic Modeling

paper · pdf · doi:10.48550/arxiv.2510.06308

openalex publication_date 2025/10/07 · openalex created_date 2025/10/18 · openalex updated_date 2026/07/28

Abstract

We introduce Lumina-DiMOO, an open-source foundational model for seamless multi-modal generation and understanding. Lumina-DiMOO sets itself apart from prior unified models by utilizing a fully discrete diffusion modeling to handle inputs and outputs across various modalities. This innovative approach allows Lumina-DiMOO to achieve higher sampling efficiency compared to previous autoregressive (AR) or hybrid AR-Diffusion paradigms and adeptly support a broad spectrum of multi-modal tasks, including text-to-image generation, image-to-image generation (e.g., image editing, subject-driven generation, and image inpainting, etc.), as well as image understanding. Lumina-DiMOO achieves state-of-the-art performance on multiple benchmarks, surpassing existing open-source unified multi-modal models. To foster further advancements in multi-modal and discrete diffusion model research, we release our code and checkpoints to the community. Project Page: https://synbol.github.io/Lumina-DiMOO.

Citations

• Pref-GRPO: Pairwise Preference Reward-based GRPO for Stable Text-to-Image Reinforcement Learning
• InternVL3.5: Advancing Open-Source Multimodal Models in Versatility, Reasoning, and Efficiency
• Lumina-mGPT 2.0: Stand-Alone AutoRegressive Image Modeling
• Resurrect Mask AutoRegressive Modeling for Efficient and Scalable Image Generation
• OmniGen2: Exploration to Advanced Multimodal Generation
• ShareGPT-4o-Image: Aligning Multimodal Models with GPT-4o-Level Image Generation
• OneIG-Bench: Omni-dimensional Nuanced Evaluation for Image Generation
• TIIF-Bench: How Does Your T2I Model Follow Your Instructions?
• UniRL: Self-Improving Unified Multimodal Models via Supervised and Reinforcement Learning
• Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding
• ImgEdit: A Unified Image Editing Dataset and Benchmark
• LLaDA 1.5: Variance-Reduced Preference Optimization for Large Language Diffusion Models
• Plug-and-Play Context Feature Reuse for Efficient Masked Generation
• Dimple: Discrete Diffusion Multimodal Large Language Model with Parallel Decoding
• LaViDa: A Large Diffusion Language Model for Multimodal Understanding
• LLaDA-V: Large Language Diffusion Models with Visual Instruction Tuning
• dKV-Cache: The Cache for Diffusion Language Models
• MMaDA: Multimodal Large Diffusion Language Models
• Emerging Properties in Unified Multimodal Pretraining
• dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Caching
• Reinforcing the Diffusion Chain of Lateral Thought with Diffusion Language Models
• BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset
• Mogao: An Omni Foundation Model for Interleaved Multi-Modal Generation
• From Reflection to Perfection: Scaling Inference-Time Optimization for Text-to-Image Diffusion Models via Reflection Tuning
• InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
• VisualCloze: A Universal Image Generation Framework via Visual In-Context Learning
• Transfer between Modalities with MetaQueries
• Lumina-OmniLV: A Unified Multimodal Framework for General Low-Level Vision
• Lumina-Image 2.0: A Unified and Efficient Image Generative Framework
• Qwen2.5-VL Technical Report
• Large Language Diffusion Models
• SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformer
• Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling
• DeepSeek-R1 incentivizes reasoning in LLMs through reinforcement learning
• Dual Diffusion for Unified Image Generation and Understanding
• Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling
• MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale
• Liquid: Language Models are Scalable and Unified Multi-modal Generators
• TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation
• MV-Adapter: Multi-view Consistent Image Generation Made Easy
• One Diffusion to Generate Them All
• AnyEdit: Mastering Unified High-Quality Image Editing for Any Idea
• OminiControl: Minimal and Universal Control for Diffusion Transformer
• JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation
• OmniEdit: Building Image Editing Generalist Models Through Specialist Supervision
• Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and Generation
• Harnessing Webpage UIs for Text-Rich Visual Understanding
• SANA: Efficient High-Resolution Image Synthesis with Linear Diffusion Transformers
• Jump Your Steps: Optimizing Sampling Schedule of Discrete Diffusion Models
• Meissonic: Revitalizing Masked Generative Transformers for Efficient High-Resolution Text-to-Image Synthesis
• Discrete Copula Diffusion
• Emu3: Next-Token Prediction is All You Need
• Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
• OmniGen: Unified Image Generation
• VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation
• Open-MAGVIT2: An Open-Source Project Toward Democratizing Auto-regressive Visual Generation
• AutoGeo: Automating Geometric Image Dataset Creation for Enhanced Geometry Understanding
• Show-o: One Single Transformer to Unify Multimodal Understanding and Generation
• Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model
• Lumina-mGPT: Illuminate Flexible Photorealistic Text-to-Image Generation with Multimodal Generative Pretraining
• The Llama 3 Herd of Models
• UltraEdit: Instruction-based Fine-Grained Image Editing at Scale
• Multimodal Table Understanding
• Simple and Effective Masked Diffusion Language Models
• Autoregressive Model Beats Diffusion: Llama for Scalable Image Generation
• Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Data
• Lumina-Next: Making Lumina-T2X Stronger and Faster with Next-DiT
• Chameleon: Mixed-Modal Early-Fusion Foundation Models
• How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites
• SEED-X: Multimodal Models with Unified Multi-granularity Comprehension and Generation
• InstantStyle: Free Lunch towards Style-Preserving in Text-to-Image Generation
• ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment
• Scaling Rectified Flow Transformers for High-Resolution Image Synthesis
• Playground v2.5: Three Insights towards Enhancing Aesthetic Quality in Text-to-Image Generation
• MobileVLM V2: Faster and Stronger Baseline for Vision Language Model
• LLaVA-Phi: Efficient Multi-Modal Assistant with Small Language Model
• aMUSEd: An Open MUSE Reproduction
• InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks
• Generative Multimodal Models are In-Context Learners
• MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI
• Davidsonian Scene Graph: Improving Reliability in Fine-grained Evaluation for Text-to-Image Generation
• Discrete Diffusion Modeling by Estimating the Ratios of the Data Distribution
• GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignment
• Improved Baselines with Visual Instruction Tuning
• Kandinsky: an Improved Text-to-Image Synthesis with Image Prior and Latent Diffusion
• PixArt-α: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis
• Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
• SEED-Bench: Benchmarking Multimodal LLMs with Generative Comprehension
• MMBench: Is Your Multi-modal Model an All-around Player?
• SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis
• Evaluating Object Hallucination in Large Vision-Language Models
• Visual Instruction Tuning
• DINOv2: Learning Robust Visual Features without Supervision
• Adding Conditional Control to Text-to-Image Diffusion Models
• Adding Conditional Control to Text-to-Image Diffusion Models
• MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessment
• MaskGIT: Masked Generative Image Transformer
• High-Resolution Image Synthesis with Latent Diffusion Models
• High-Resolution Image Synthesis with Latent Diffusion Models
• LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs
• MUSIQ: Multi-scale Image Quality Transformer
• Structured Denoising Diffusion Models in Discrete State-Spaces
• Learning Transferable Visual Models From Natural Language Supervision
• Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize\n Long-Tail Visual Concepts

Cited by

• dMLLM-TTS: Self-Verified and Efficient Test-Time Scaling for Diffusion Multi-Modal Large Language Models
• UmniBench: Unified Understand and Generation Model Oriented Omni-dimensional Benchmark
• Multi-Agent Medical Decision Consensus Matrix System: An Intelligent Collaborative Framework for Oncology MDT Consultations
• ViewMask-1-to-3: Multi-View Consistent Image Generation via Multimodal Discrete Diffusion Models
• SDAR-VL: Stable and Efficient Block-wise Diffusion for Vision-Language Understanding
• RecTok: Reconstruction Distillation along Rectified Flow
• Adaptive Detector-Verifier Framework for Zero-Shot Polyp Detection in Open-World Settings
• DraCo: Draft as CoT for Text-to-Image Preview and Rare Concept Generation
• HUD: Hierarchical Uncertainty-Aware Disambiguation Network for Composed Video Retrieval
• Does Understanding Inform Generation in Unified Multimodal Models? From Analysis to Path Forward
• Calibrated Multimodal Representation Learning with Missing Modalities
• MMaDA-Parallel: Multimodal Large Diffusion Language Models for Thinking-Aware Editing and Generation
• UniREditBench: A Unified Reasoning-based Image Editing Benchmark
• Thinking with Video: Video Generation as a Promising Multimodal Reasoning Paradigm
• UniMedVL: Unifying Medical Multimodal Understanding And Generation Through Observation-Knowledge-Analysis
• From Masks to Worlds: A Hitchhiker's Guide to World Models
• A Survey on Cache Methods in Diffusion Models: Toward Efficient Multi-Modal Generation
• UniGenBench++: A Unified Semantic Evaluation Benchmark for Text-to-Image Generation
• PICABench: How Far Are We from Physically Realistic Image Editing?
• Planning with Unified Multimodal Models
• RLGF: Reinforcement Learning with Geometric Feedback for Autonomous Driving Video Generation

Related

• Unified Multimodal Understanding and Generation Models: Advances, Challenges, and Opportunities
• M-VADER: A Model for Diffusion with Multimodal Context
• Lavida-O: Elastic Large Masked Diffusion Models for Unified Multimodal Understanding and Generation
• DiffBlender: Composable and Versatile Multimodal Text-to-Image Diffusion Models
• OmniGen: Unified Image Generation