mlsanityData to 14 Oct 2026

NextStep-1: Toward Autoregressive Image Generation with Continuous Tokens at Scale

arXiv 2508.10711 ↗First posted 14 Aug 2025cs.CVcited by 57

On the map: Diffusion models → Image generation, text-to-image and editing

Abstract (arXiv)

Prevailing autoregressive (AR) models for text-to-image generation either rely on heavy, computationally-intensive diffusion models to process continuous image tokens, or employ vector quantization (VQ) to obtain discrete tokens with quantization loss. In this paper, we push the autoregressive paradigm forward with NextStep-1, a 14B autoregressive model paired with a 157M flow matching head, training on discrete text tokens and continuous image tokens with next-token prediction objectives. NextStep-1 achieves state-of-the-art performance for autoregressive models in text-to-image generation tasks, exhibiting strong capabilities in high-fidelity image synthesis. Furthermore, our method shows strong performance in image editing, highlighting the power and versatility of our unified approach. To facilitate open research, we will release our code and models to the community.

Key facts. Cited by 57 papers in the map (data through 14 Oct 2026). Placed in Image generation, text-to-image and editing, one of 14 directions in Diffusion models. Builds on 83 of its 97 references that are in the map. First posted on arXiv 14 Aug 2025.

Builds on

83 of 97 references.

Closest ancestors · 12

Same direction, newest first

  1. 2025Show-o2: Improved Native Unified Multimodal Models18 Jun
  2. 2025FLUX.1 Kontext: Flow Matching for In-Context Image Generation and Editing in Latent Space17 Jun
  3. 2025BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset14 May
  4. 2024SeedEdit: Align Image Re-Generation to Image Editing11 Nov
  5. 2024Emu3: Next-Token Prediction is All You Need27 Sep
  6. 2024OmniGen: Unified Image Generation17 Sep
  7. 2024Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction3 Apr
  8. 2024ELLA: Equip Diffusion Models with LLM for Enhanced Semantic Alignment8 Mar
  9. 2024Playground v2.5: Three Insights towards Enhancing Aesthetic Quality in Text-to-Image Generation27 Feb
  10. 2023GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignment17 Oct
  11. 2023Adding Conditional Control to Text-to-Image Diffusion Models10 Feb
  12. 2022InstructPix2Pix: Learning to Follow Image Editing Instructions17 Nov

Other references · 71

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,863
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  4. 2023Segment Anythingcited by 8,234
  5. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 8,065
  6. 2022Flow Matching for Generative Modelingcited by 5,702
  7. 2025Qwen2.5-VL Technical Reportcited by 5,587
  8. 2024Qwen2.5 Technical Reportcited by 4,942
  9. 2021RoFormer: Enhanced Transformer with Rotary Position Embeddingcited by 4,703
  10. 2023SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesiscited by 4,356
  11. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  12. 2024Scaling Rectified Flow Transformers for High-Resolution Image Synthesiscited by 4,181
38 more
  1. 2020Taming Transformers for High-Resolution Image Synthesiscited by 3,253
  2. 2022DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Stepscited by 1,979
  3. 2022Fast Inference from Transformers via Speculative Decodingcited by 1,653
  4. 2022Scaling Autoregressive Models for Content-Rich Text-to-Image Generationcited by 1,295
  5. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,045
  6. 2025Qwen-Image Technical Reportcited by 1,044
  7. 2022DPM-Solver++: Fast Solver for Guided Sampling of Diffusion Probabilistic Modelscited by 938
  8. 2025Emerging Properties in Unified Multimodal Pretrainingcited by 897
  9. 2025Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scalingcited by 812
  10. 2023Diffusion Model Alignment Using Direct Preference Optimizationcited by 790
  11. 2024Autoregressive Model Beats Diffusion: Llama for Scalable Image Generationcited by 777
  12. 2023Objaverse-XL: A Universe of 10M+ 3D Objectscited by 774
  13. 2024Show-o: One Single Transformer to Unify Multimodal Understanding and Generationcited by 752
  14. 2022On Distillation of Guided Diffusion Modelscited by 725
  15. 2024Diffusion Forcing: Next-token Prediction Meets Full-Sequence Diffusioncited by 681
  16. 2024Autoregressive Image Generation without Vector Quantizationcited by 651
  17. 2023Language Model Beats Diffusion -- Tokenizer is Key to Visual Generationcited by 640
  18. 2023MagicBrush: A Manually Annotated Dataset for Instruction-Guided Image Editingcited by 624
  19. 2024Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Modelcited by 483
  20. 2024Evaluating Text-to-Visual Generation with Image-to-Text Generationcited by 474
  21. 2023Generative Multimodal Models are In-Context Learnerscited by 446
  22. 2025Step1X-Edit: A Practical Framework for General Image Editingcited by 429
  23. 2025Reconstruction vs. Generation: Taming Optimization Dilemma in Latent Diffusion Modelscited by 393
  24. 2024Better & Faster Large Language Models via Multi-token Predictioncited by 360
  25. 2024SEED-X: Multimodal Models with Unified Multi-granularity Comprehension and Generationcited by 331
  26. 2024Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesiscited by 304
  27. 2024An Image is Worth 32 Tokens for Reconstruction and Generationcited by 302
  28. 2023DreamLLM: Synergistic Multimodal Comprehension and Creationcited by 297
  29. 2023Emu: Enhancing Image Generation Models Using Photogenic Needles in a Haystackcited by 279
  30. 2025OmniGen2: Towards Instruction-Aligned Multimodal Generationcited by 253
  31. 2024AnyEdit: Mastering Unified High-Quality Image Editing for Any Ideacited by 241
  32. 2025Transfer between Modalities with MetaQueriescited by 236
  33. 2025ImgEdit: A Unified Image Editing Dataset and Benchmarkcited by 231
  34. 2024MetaMorph: Multimodal Understanding and Generation via Instruction Tuningcited by 206
  35. 2025WISE: A World Knowledge-Informed Semantic Evaluation for Text-to-Image Generationcited by 204
  36. 2025Step-Video-T2V Technical Report: The Practice, Challenges, and Future of Video Foundation Modelcited by 168
  37. 2025SANA 1.5: Efficient Scaling of Training-Time and Inference-Time Compute in Linear Diffusion Transformercited by 165
  38. 2025Seedream 3.0 Technical Reportcited by 154

50 most cited of 71 All 71 (JSON)

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 20; 2026: 37.

57 papers.

Sort:

By area:

  1. 2025InfinityStar: Unified Spacetime AutoRegressive Modeling for Visual Generationcited by 45
  2. 2026ImageWAM: Do World Action Models Really Need Video Generation, or Just Image Editing?cited by 38
  3. 2025ReasonEdit: Towards Reasoning-Enhanced Image Editing Modelscited by 25
  4. 2025X-Streamer: Unified Human World Modeling with Audiovisual Interactioncited by 18
  5. 2025Uni-X: Mitigating Modality Conflict with a Two-End-Separated Architecture for Unified Multimodal Modelscited by 18
  6. 2025FARMER: Flow AutoRegressive Transformer over Pixelscited by 18
  7. 2025Hyperspherical Latents Improve Continuous-Token Autoregressive Generationcited by 16
  8. 2026Everything in Its Place: Benchmarking Spatial Intelligence of Text-to-Image Modelscited by 16
  9. 2026NextFlow: Unified Sequential Modeling Activates Multimodal Understanding and Generationcited by 14
  10. 2026GENIUS: Generative Fluid Intelligence Evaluation Suitecited by 12
  11. 2026LLaMo: Scaling Pretrained Language Models for Unified Motion Understanding and Generation with Continuous Autoregressive Tokenscited by 11
  12. 2025Continuous Autoregressive Language Modelscited by 10
38 more
  1. 2025VQRAE: Representation Quantization Autoencoders for Multimodal Understanding, Generation and Reconstructioncited by 10
  2. 2025UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Editscited by 10
  3. 2025SimFlow: Simplified and End-to-End Training of Latent Normalizing Flowscited by 7
  4. 2026Spectrum Matching: a Unified Perspective for Superior Diffusability in Latent Diffusioncited by 7
  5. 2026From Broad Exploration to Stable Synthesis: Entropy-Guided Optimization for Autoregressive Image Generationcited by 7
  6. 2026Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuningcited by 6
  7. 2025UniRL-Zero: Reinforcement Learning on Unified Models with Joint Language Model and Diffusion Model Expertscited by 5
  8. 2026GUI-GenBench: Evaluating Image Generation Models as Interactive GUI Environmentscited by 5
  9. 2026Visual Generation in the New Era: An Evolution from Atomic Mapping to Agentic World Modelingcited by 5
  10. 2025Direct Preference Optimization for Speech Autoregressive Diffusion Modelscited by 4
  11. 2025Exposing Blindspots: Cultural Bias Evaluation in Generative Image Modelscited by 4
  12. 2026PlanViz: Evaluating Planning-Oriented Image Generation and Editing for Computer-Use Taskscited by 4
  13. 2026BitDance: Scaling Autoregressive Generative Models with Binary Tokenscited by 4
  14. 2025Autoregressive Styled Text Image Generation, but Make it Reliablecited by 3
  15. 2026Generative Refinement Networks for Visual Synthesiscited by 3
  16. 2026Steering Visual Generation in Unified Multimodal Models with Understanding Supervisioncited by 3
  17. 2026FlashAR: Efficient Post-Training Acceleration for Autoregressive Image Generationcited by 3
  18. 2025Visual Generation Tuningcited by 2
  19. 2025Fast-ARDiff: An Entropy-informed Acceleration Framework for Continuous Space Autoregressive Generationcited by 2
  20. 2025ARMFlow: AutoRegressive MeanFlow for Online 3D Human Reaction Generationcited by 2
  21. 2026MAR-GRPO: Stabilized GRPO for AR-diffusion Hybrid Image Generationcited by 2
  22. 2026Uni-ViGU: Towards Unified Video Generation and Understanding via A Diffusion-Based Video Generatorcited by 2
  23. 2025Optimizing Input of Denoising Score Matching is Biased Towards Higher Score Normcited by 1
  24. 2025Diversifying Long Prompt Image Generation through Structured Prompt Embedding Space Samplingcited by 1
  25. 2026Shifting the Breaking Point of Flow Matching for Multi-Instance Editingcited by 1
  26. 2026Semantic-Aware Prefix Learning for Token-Efficient Image Generationcited by 1
  27. 2026On the Robustness of Watermarking for Autoregressive Image Generationcited by 1
  28. 2026VibeToken: Scaling 1D Image Tokenizers and Autoregressive Models for Dynamic Resolution Generationscited by 1
  29. 2026F3-Tokenizer: Taming Audio Autoencoder Latents for Understanding and Generationcited by 1
  30. 2026Autoregressive Boltzmann Generatorscited by 1
  31. 2026KVAE: Family of Tokenizers for Multimodal Generative Modelscited by 1
  32. 2025Exploring the Design Space of Transition Matchingcited by 0
  33. 2026LINA: Linear Autoregressive Image Generative Models with Continuous Tokenscited by 0
  34. 2026UniWeTok: An Unified Binary Tokenizer with Codebook Size mathit2128 for Unified Multimodal Large Language Modelcited by 0
  35. 2026From "What" to "How": Constrained Reasoning for Autoregressive Image Generationcited by 0
  36. 2026Drift-AR: Single-Step Visual Autoregressive Generation via Anti-Symmetric Driftingcited by 0
  37. 2026DIVA: Harnessing the Representation Divergence in Unified Multimodal Models for Mutual Reinforcementcited by 0
  38. 2026Channel-wise Vector Quantizationcited by 0

50 most cited of 57 All 57 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2025Fast Autoregressive Models for Continuous Latent Generation
  2. 2023Emage: Non-Autoregressive Text-to-Image Generation
  3. 2022Progressive Text-to-Image Generation
  4. 2025Frequency Autoregressive Image Generation with Continuous Tokens
  5. 2025EditAR: Unified Conditional Generation with Autoregressive Models
  6. 2025Improving Autoregressive Image Generation through Coarse-to-Fine Token Prediction
  7. 2025Beyond Next-Token: Next-X Prediction for Autoregressive Visual Generation
  8. 2024FlowAR: Scale-wise Autoregressive Image Generation Meets Flow Matching