mlsanityData to 7 Oct 2026

Compute Only 16 Tokens in One Timestep: Accelerating Diffusion Transformers with Cluster-Driven Feature Caching

arXiv 2509.10312 ↗First posted 12 Sep 2025cs.CVcited by 17

On the map: Diffusion models → Diffusion foundations and sampling

Abstract (arXiv)

Diffusion transformers have gained significant attention in recent years for their ability to generate high-quality images and videos, yet still suffer from a huge computational cost due to their iterative denoising process. Recently, feature caching has been introduced to accelerate diffusion transformers by caching the feature computation in previous timesteps and reusing it in the following timesteps, which leverage the temporal similarity of diffusion models while ignoring the similarity in the spatial dimension. In this paper, we introduce Cluster-Driven Feature Caching (ClusCa) as an orthogonal and complementary perspective for previous feature caching. Specifically, ClusCa performs spatial clustering on tokens in each timestep, computes only one token in each cluster and propagates their information to all the other tokens, which is able to reduce the number of tokens by over 90%. Extensive experiments on DiT, FLUX and HunyuanVideo demonstrate its effectiveness in both text-to-image and text-to-video generation. Besides, it can be directly applied to any diffusion transformer without requirements for training. For instance, ClusCa achieves 4.96x acceleration on FLUX with an ImageReward of 99.49%, surpassing the original model by 0.51%. The code is available at https://github.com/Shenyi-Z/Cache4Diffusion.

Key facts. Cited by 17 papers in the map (data through 7 Oct 2026). Placed in Diffusion foundations and sampling, one of 14 directions in Diffusion models. Builds on 27 of its 36 references that are in the map. First posted on arXiv 12 Sep 2025.

Builds on

27 of 36 references.

Closest ancestors · 15

Same direction, newest first

  1. 2025From Reusing to Forecasting: Accelerating Diffusion Models with TaylorSeers10 Mar
  2. 2025Ditto: Accelerating Diffusion Model via Temporal Value Similarity20 Jan
  3. 2025Cached Adaptive Token Merging: Dynamic Token Reduction and Redundant Computation Elimination in Diffusion Model1 Jan
  4. 2024Token Pruning for Caching Better: 9 Times Acceleration on Stable Diffusion for Free31 Dec
  5. 2024Rethinking Token-wise Feature Caching: Accelerating Diffusion Transformers with Dual Feature Caching25 Dec
  6. 2024DiP-GO: A Diffusion Pruner via Few-step Gradient Optimization22 Oct
  7. 2024Accelerating Diffusion Transformers with Token-wise Feature Caching5 Oct
  8. 2024FORA: Fast-Forward Caching in Diffusion Transformer Acceleration1 Jul
  9. 2024Delta-DiT: A Training-Free Acceleration Method Tailored for Diffusion Transformers3 Jun
  10. 2023DeepCache: Accelerating Diffusion Models for Free1 Dec
  11. 2023Token Merging for Fast Stable Diffusion30 Mar
  12. 2022Post-training Quantization on Diffusion Models28 Nov
3 more
  1. 2022DPM-Solver: A Fast ODE Solver for Diffusion Probabilistic Model Sampling in Around 10 Steps2 Jun
  2. 2021Diffusion Models Beat GANs on Image Synthesis11 May
  3. 2020Denoising Diffusion Implicit Models6 Oct

Other references · 12

  1. 2014ImageNet Large Scale Visual Recognition Challengecited by 8,441
  2. 2022Photorealistic Text-to-Image Diffusion Models with Deep Language Understandingcited by 6,322
  3. 2022FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awarenesscited by 3,611
  4. 2022Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flowcited by 3,550
  5. 2023Stable Video Diffusion: Scaling Latent Video Diffusion Models to Large Datasetscited by 2,526
  6. 2023VBench: Comprehensive Benchmark Suite for Video Generative Modelscited by 1,755
  7. 2023ImageReward: Learning and Evaluating Human Preferences for Text-to-Image Generationcited by 1,010
  8. 2024Timestep Embedding Tells: It's Time to Cache for Video Diffusion Modelcited by 245
  9. 2024Real-Time Video Generation with Pyramid Attention Broadcastcited by 140
  10. 2023Token Fusion: Bridging the Gap between Token Pruning and Token Mergingcited by 123
  11. 2024DiTFastAttn: Attention Compression for Diffusion Transformer Modelscited by 109
  12. 2024Hunyuan-Large: An Open-Source MoE Model with 52 Billion Activated Parameters by Tencentcited by 89

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 15.

17 papers.

Sort:

By area:

  1. 2026LiveEdit: Towards Real-Time Diffusion-Based Streaming Video Editingcited by 15
  2. 2025A Survey on Cache Methods in Diffusion Models: Toward Efficient Multi-Modal Generationcited by 14
  3. 2026Adaptive Spectral Feature Forecasting for Diffusion Sampling Accelerationcited by 5
  4. 2026DisCa: Accelerating Video Diffusion Transformers with Distillation-Compatible Learnable Feature Cachingcited by 4
  5. 2026LESA: Learnable Stage-Aware Predictors for Diffusion Model Accelerationcited by 4
  6. 2026WorldCache: Accelerating World Models for Free via Heterogeneous Token Cachingcited by 4
  7. 2026ResilPhase: Plug-and-Play Phase Mapping and Noise-Resilient Macro-Trajectory Extrapolation for Diffusion Accelerationcited by 3
  8. 2026Motion-Aware Caching for Efficient Autoregressive Video Generationcited by 2
  9. 2025Generative AI Meets 6G and Beyond: Diffusion Models for Semantic Communicationscited by 1
  10. 2026BeCARE: Budgeted Cache Refresh for Diffusion Transformer Accelerationcited by 1
  11. 2026SAFE-DiT: Semantics-Aware Fast-path Execution for High-Resolution Diffusion Transformerscited by 1
  12. 2026OnlineCache: Learning Dynamic Caching Policies with Error Correction for Efficient Diffusion Inferencecited by 1
5 more
  1. 2026Disagree to Accelerate: Closing the Loop on Diffusion Feature Forecastscited by 1
  2. 2026SyncCache: Exploiting Asymmetric Dynamics for Fast Audio-Driven Portrait Animationcited by 0
  3. 2026LinCa: Accelerating Diffusion Models via Learnable Decomposed Feature Cachingcited by 0
  4. 2026Beyond Visual Similarity: Entity-Aligned Retrieval for Knowledge-Based Visual Question Answeringcited by 0
  5. 2026Accelerating Diffusion Transformers with Gaussian Process Rectified Feature Cachecited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Let Features Decide Their Own Solvers: Hybrid Feature Caching for Diffusion Transformers
  2. 2025ProCache: Constraint-Aware Feature Caching with Selective Computation for Diffusion Transformer Acceleration
  3. 2025BWCache: Accelerating Video Diffusion Transformers through Block-Wise Caching
  4. 2025FreqCa: Accelerating Diffusion Models via Frequency-Aware Caching
  5. 2025Forecast then Calibrate: Feature Caching as ODE for Efficient Diffusion Transformers
  6. 2026Forecast the Principal, Stabilize the Residual: Subspace-Aware Feature Caching for Efficient Diffusion Transformers
  7. 2026OmniCache: Multidimensional Hierarchical Feature Caching For Diffusion Models
  8. 2025HiCache: A Plug-in Scaled-Hermite Upgrade for Taylor-Style Cache-then-Forecast Diffusion Acceleration