Lip Forcing distills a 14B bidirectional video diffusion teacher into autoregressive students that achieve real-time lip synchronization at 31 FPS using two denoising steps without CFG.
super hub Canonical reference
HunyuanVideo: A Systematic Framework For Large Video Generative Models
Canonical reference. 85% of citing Pith papers cite this work as background.
364
Pith papers citing it
Background
85% of classified citations
abstract
Recent advancements in video generation have significantly impacted daily life for both individuals and industries. However, the leading video generation models remain closed-source, resulting in a notable performance gap between industry capabilities and those available to the public. In this report, we introduce HunyuanVideo, an innovative open-source video foundation model that demonstrates performance in video generation comparable to, or even surpassing, that of leading closed-source models. HunyuanVideo encompasses a comprehensive framework that integrates several key elements, including data curation, advanced architectural design, progressive model scaling and training, and an efficient infrastructure tailored for large-scale model training and inference. As a result, we successfully trained a video generative model with over 13 billion parameters, making it the largest among all open-source models. We conducted extensive experiments and implemented a series of targeted designs to ensure high visual quality, motion dynamics, text-video alignment, and advanced filming techniques. According to evaluations by professionals, HunyuanVideo outperforms previous state-of-the-art models, including Runway Gen-3, Luma 1.6, and three top-performing Chinese video generative models. By releasing the code for the foundation model and its applications, we aim to bridge the gap between closed-source and open-source communities. This initiative will empower individuals within the community to experiment with their ideas, fostering a more dynamic and vibrant video generation ecosystem. The code is publicly available at https://github.com/Tencent/HunyuanVideo.
hub tools
citation-role summary
background 73
baseline 6
method 3
dataset 2
other 1
citation-polarity summary
claims ledger
- abstract Recent advancements in video generation have significantly impacted daily life for both individuals and industries. However, the leading video generation models remain closed-source, resulting in a notable performance gap between industry capabilities and those available to the public. In this report, we introduce HunyuanVideo, an innovative open-source video foundation model that demonstrates performance in video generation comparable to, or even surpassing, that of leading closed-source models. HunyuanVideo encompasses a comprehensive framework that integrates several key elements, including
authors
co-cited works
representative citing papers
AnyFlow enables any-step video diffusion by distilling flow-map transitions over arbitrary time intervals with on-policy backward simulation.
TrackCraft3R is the first method to repurpose a video diffusion transformer as a feed-forward dense 3D tracker via dual-latent representations and temporal RoPE alignment, achieving SOTA performance with lower compute.
Flow-GRPO is the first online RL method for flow matching models, raising GenEval accuracy from 63% to 95% and text-rendering accuracy from 59% to 92% with little reward hacking.
OrbitQuant is a data-agnostic PTQ technique for DiTs that uses RPBH rotation in a normalized basis to enable a single codebook across all inputs, achieving SOTA low-bit performance on FLUX.1, CogVideoX and similar models.
NEvo performs evolutionary search guided by a dynamic voxel-level encoding model to synthesize videos that maximize predicted activity in target brain ROIs, recovering known selectivities and revealing temporal dynamics differences.
QWERTY enables training-free motion control in pretrained image-to-video DiTs by warping the frame-invariant semantic subspace of queries in 3D full attention and using the predicted noise as self-guidance for latent optimization.
Introduces VidPair-Halluc benchmark of 1K background-controlled adversarial video pairs and 11K QA pairs generated via PairFlow pipeline to evaluate hallucination in LVMs.
MemLearner introduces a learning-based adaptive context query method using query tokens in video world models to improve long-term scene consistency over rule-based retrieval.
Introduces CIPE-Dance as the largest dance video dataset and OmniDance framework for unified text-music multimodal dance video generation achieving SOTA on TI2V, MI2V, and MTI2V tasks.
Uni-Mo generates 7,488 language-annotated quadruped motions via LLM prompts and video diffusion, lifts them to 3D trajectories, and trains policies achieving 96.7% real-robot success on 392 sampled motions.
RayPE extends video DiT attention with Plucker coordinates and a gated reciprocal-product term to improve 3D consistency and camera controllability.
DRIVE-CHOREO uses three LLM agents to create a unified position-aware token sequence co-compressed with multi-view video, achieving SOTA BEV mAP of 21.6 and +2.4 NDS improvement on nuScenes.
CineOrchestra unifies control of subjects, events, cameras, and shot transitions in cinematic video generation through entity-centric conditioning primitives and parameter-free coordinated rotary embeddings.
Self-distillation from a caption-conditioned video diffusion model to an image-and-prompt-conditioned executor, enhanced by RL from VLM feedback, enables task solving in world models.
FadeMem introduces distance-aware KV memory consolidation for autoregressive video diffusion that builds a temporal hierarchy with power-law merging to preserve short-term dynamics and long-range coherence under fixed cache budget.
OmniTryOn performs multi-object video virtual try-on in one pass using first-frame wearable caching and spatiotemporal RoPE, outperforming single-garment baselines on a new TryAny-Bench dataset.
Dream.exe evaluates 8 video generation models on 101 manipulation tasks by converting generated videos into executable robot trajectories in a simulator, finding measurable success rates that visual metrics do not predict.
LA-LQR applies latent-space linear-quadratic regulator control to steer text-to-video model activations toward desired features while penalizing excessive changes.
DCVC-UF uses chunk-based joint encoding and parallel frame-specific decoding to deliver ultra-fast neural video compression while claiming new state-of-the-art rate-distortion performance.
A large-scale empirical study across tokenizers and diffusion backbones identifies Velocity Irreducible Variance (VIV) as one of the most stable predictors of latent diffusion generation quality.
SPAWN enables training-free insertion of custom visual concepts into autoregressive world models by swapping the pinned context-memory anchor over a short injection window.
VLMs formulate differentiable rewards from task-specific rules to enable test-time online LoRA optimization of VGMs, delivering 16.7-point gains on symbolic and general video reasoning benchmarks over VLM-as-solver and Best-of-N baselines.
LongLive-RAG formulates long video generation as retrieval-augmented generation by treating self-generated latents as a dynamic searchable history and adding a Window Temporal Delta Loss for better retrieval.
citing papers explorer
-
LongLive: Real-time Interactive Long Video Generation
LongLive is a causal autoregressive video generator that produces up to 240-second interactive videos at 20.7 FPS on one H100 GPU after 32 GPU-days of fine-tuning from a 1.3B short-clip model.
-
EditVerse: Unifying Image and Video Editing and Generation with In-Context Learning
EditVerse unifies image and video editing and generation in one transformer model via unified token sequences and in-context learning, trained jointly on curated video editing data plus image/video corpora and evaluated on a new instruction-based benchmark.
-
HERO: Hierarchical Extrapolation and Refresh for Efficient World Models
HERO accelerates world model inference 1.73x via hierarchical patch-wise refresh in shallow layers and linear extrapolation in deeper layers with minimal quality loss.
-
Vidar: Embodied Video Diffusion Model for Generalist Manipulation
Vidar shows that a video diffusion prior continuously pre-trained on 750K multi-view robot trajectories plus a label-free masked inverse dynamics adapter can generalize manipulation to new robot embodiments with 1% of typical demonstration data.
-
Geometry Forcing: Marrying Video Diffusion and 3D Representation for Consistent World Modeling
Geometry Forcing aligns video diffusion representations with geometric foundation model features via angular cosine and scale regression objectives to improve 3D consistency in generated videos.
-
Listener-Rewarded Thinking in VLMs for Image Preferences
Listener-augmented GRPO uses an independent frozen VLM to provide dense confidence scores on reasoning traces, yielding 67.4% accuracy on ImageReward, up to +6% OOD gains on 1.2M-vote human data, and fewer reasoning contradictions.
-
Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion
Self Forcing trains autoregressive video diffusion models by performing autoregressive rollout with KV caching during training to close the exposure bias gap, using a holistic video-level loss and few-step diffusion for efficiency.
-
LLaDA-V: Large Language Diffusion Models with Visual Instruction Tuning
LLaDA-V is a diffusion-based multimodal large language model that reaches competitive or state-of-the-art results on visual instruction tasks while using a non-autoregressive architecture.
-
MAGI-1: Autoregressive Video Generation at Scale
MAGI-1 is a 24B-parameter autoregressive video world model that predicts denoised frame chunks sequentially with increasing noise to enable causal, scalable, streaming generation up to 4M token contexts.
-
DanceGRPO: Unleashing GRPO on Visual Generation
DanceGRPO applies GRPO to visual generation tasks to achieve stable policy optimization across diffusion models, rectified flows, multiple tasks, and diverse reward models, outperforming prior RL methods.
-
We'll Fix it in Post: Improving Text-to-Video Generation with Neuro-Symbolic Feedback
NeuS-E is a post-generation refinement method that uses neuro-symbolic analysis of a formal video representation to detect and correct semantic and temporal inconsistencies in text-to-video outputs, improving prompt alignment by nearly 40%.
-
Long-Context Autoregressive Video Modeling with Next-Frame Prediction
FAR baseline plus asymmetric kernels for long short-term context modeling achieves SOTA short and long video generation in autoregressive setups.
-
Improving Video Generation with Human Feedback
A human preference dataset and VideoReward model enable Flow-DPO and Flow-NRG to produce smoother, better-aligned videos from text prompts in flow-based generators.
-
LTX-Video: Realtime Video Latent Diffusion
LTX-Video integrates Video-VAE and transformer for 1:192 latent compression and real-time video diffusion by moving patchifying to the VAE and letting the decoder finish denoising in pixel space.
-
PipeFusion: Patch-level Pipeline Parallelism for Diffusion Transformers Inference
PipeFusion applies patch partitioning and pipeline parallelism with one-step stale feature reuse to reduce communication overhead in DiT inference, reporting SOTA results on 8x L40 GPUs for Pixart, SD3, and Flux.1.
-
Latte: Latent Diffusion Transformer for Video Generation
Latte achieves state-of-the-art video generation on FaceForensics, SkyTimelapse, UCF101, and Taichi-HD by using a latent diffusion transformer with four efficient spatial-temporal decomposition variants and best-practice training choices.
-
Stream-T1: Test-Time Scaling for Streaming Video Generation
Stream-T1 is a test-time scaling framework for streaming video generation using scaled noise propagation from history, reward pruning across short and long windows, and feedback-guided memory sinking to improve temporal consistency and visual quality.
-
PhysMani: Physics-principled 3D World Model for Dynamic Object Manipulation
PhysMani couples a physics-principled 3D Gaussian world model with a future-aware policy to achieve higher success rates on dynamic manipulation tasks in simulation and real robots.
-
Arachne: Orchestrating Cascades for Efficient Text-to-Video Model Training
Arachne orchestrates cascades for distributed T2V training and reports up to 65% lower iteration time with improving gains at larger scales compared to static bucketing approaches.
-
RetailSMV: Exocentric vs. Egocentric Adaptation of Foundation Video World Models in Retail
Exocentric-only LoRA adaptation of Cosmos3-Nano on a new synchronized retail video dataset matches or exceeds combined ego+exo training on most held-out metrics.
-
Bridging Video Understanding and Generation in a Unified Framework
Vega unifies video understanding and generation via shared vocabulary and hybrid autoregressive-diffusion architecture, reporting strong results on VBench and VideoMME.
-
OTCache: Optimal Transport for Geometry-Aware Caching in Diffusion Models
OTCache uses optimal transport to interpolate caching schedules between a graph-based reference and an Optuna-optimized anchor, delivering 3.66x-4.7x speedups on FLUX.1, Qwen-Image and HunyuanVideo with improved fidelity.
-
PhysisForcing: Physics Reinforced World Simulator for Robotic Manipulation
PhysisForcing applies trajectory and relational alignment losses to DiT features in video models, improving physical plausibility on R-Bench, PAI-Bench, and EZS-Bench while raising closed-loop robotic success rates from 16% to 24%.
-
Data-Forcing Distillation: Restoring Diversity and Fidelity in Few-Step Video Generation
Data-Forcing Distillation adds a teacher score discrepancy term to DMD-style distillation, restoring diversity and fidelity in few-step video models with 100-300 finetuning steps.
-
PAIWorld: A 3D-Consistent World Foundation Model for Robotic Manipulation
PAIWorld adds explicit geometric cross-view mechanisms and 3D distillation to DiT world models to achieve multi-view 3D consistency in robotic manipulation benchmarks.
-
BrainWorld: A Structural-Prior-Conditioned Generative Model for Whole-Brain 4D fMRI Dynamics
BrainWorld is a structural-prior-conditioned generative model that produces stable whole-brain 4D fMRI trajectories up to 400 frames, augments downstream tasks, and learns transferable multimodal representations across 22 datasets.
-
Kairos: A Native World Model Stack for Physical AI
Kairos is a native world model stack using cross-embodiment pretraining, hybrid linear temporal attention with theoretical error bounds, and deployment-aware co-design, reporting top performance on embodied benchmarks.
-
WorldOlympiad: Can Your World Model Survive a Triathlon?
WorldOlympiad is a new benchmark decomposing world-model evaluation into physical, geometry, and interaction tracks using segmentation, MLLM judges, Gaussian splatting, and action prompts on diverse scenarios.
-
CineDance: Towards Next-Generation Multi-Shot Long-Form Cinematic Audio-Video Generation
Introduces CineDance-1M dataset for multi-shot long-form text-to-audio-video generation along with CineBench and a model adaptation.
-
Ultra Flash: Scaling Real-Time Streaming Video Generation to High Resolutions
Ultra Flash introduces a cascaded streaming super-resolution framework with specialized training, upsampling, and optimization to enable real-time high-resolution video generation from low-res diffusion models.
-
MilliVid: Hierarchical Latents for Long-Range Consistency in Video Generation
MilliVid compresses video frames into multi-scale token hierarchies and uses coarse-to-fine rollout in a diffusion model to maintain long-range geometric and object consistency on Minecraft videos.
-
AdaGRPO: A Capability-Aware Adaptive Enhancement for Flow-based GRPO
AdaGRPO enhances GRPO for flow models via online curriculum filtering of prompts and cross-level advantage fusion, yielding performance gains and training stability.
-
Let It Be Simple: One-Step Action Generation for Vision-Language-Action Models
Biasing the training time distribution toward high-noise states enables one-step action generation in VLA models that matches or exceeds ten-step decoding on LIBERO benchmarks and real-robot tasks.
-
Physics-Informed Video Generation via Mixture-of-Experts Latent Alignment
PILA aligns frozen flow-matching video models to a physics attribute bank via MoE experts and operational residuals, reporting SOTA physical plausibility on VBench-2.0, VideoPhy-2 and PhyGenBench while preserving visual quality.
-
$\tau_0$-WM: A Unified Video-Action World Model for Robotic Manipulation
A shared video diffusion backbone jointly predicts future latents and continuous actions while also rolling out candidate actions to predict dense task-progress scores, trained on 27,300 hours of mixed robot and human data.
-
OptiWorld: Optimal Control for Video World Generation under Physical Constraints
OptiWorld inserts a classical optimal-control layer that extracts a world state, plans an optimal trajectory on a geometric manifold under physical constraints, and renders the video conditioned on that trajectory.
-
Real2SAM2Real: Generative 3D Caches as Complementary Context for Video Diffusion
Real2SAM2Real uses 3D caches from lifting models as complementary context for video diffusion models to enable precise decoupled control over camera trajectories and multi-entity motions while maintaining spatiotemporal consistency.
-
DecMem: Towards Minute-Long Consistent World Generation with Decoupled Memory
DecMem proposes a decoupled memory system using sparse global and anchored local components to enable consistent minute-long controllable video generation in world models.
-
GenEraser: Generalizable Video Object Removal via Balanced Text-Mask Guidance and Decoupled Locator-Preserver
GenEraser proposes MC-MoE with bipartite text guidance, LD-CFG fusion, and a decoupled locator-preserver architecture for generalizable video object and effect removal, claiming 2.16 dB and 1.44 dB gains on ROSE and VOR-Eval benchmarks.
-
KGEdit: Ambiguity-Aware Knowledge Graphs for Training-Free Precise Video Generation and Editing
KGEdit uses an ambiguity-aware knowledge graph and structured injection modules to improve semantic control and temporal consistency in training-free text-to-video diffusion models.
-
PixelWizard: Towards Efficient High-Fidelity Video Generation at Ultra-Large Spatial Resolution
PixelWizard decouples global structure from fine details via a spatiotemporal anchor and introduces Noise-Span Aligned Shortcut Training with biased sampling to achieve over 10x faster sampling for high-fidelity 2K/4K video generation.
-
WorldCraft: From Camera Navigation to Object Manipulation in Interactive Video World Models
WorldCraft introduces NWT, SP-LoRA, and TASP to enable object trajectory control in video-based world models while preserving camera navigation.
-
Tempered Self-Similarity Alignment for Physically Plausible Video Generation
Tempered Self-similarity Alignment transfers relational structure from foundation-model STSS into video generators via probabilistic correspondence alignment, yielding reported gains in physical plausibility on VideoPhy benchmarks.
-
Precise: SDE-Consistent Stochastic Sampling for RL Post-Training of Flow-Matching Models
Precise is a new SDE-consistent stochastic sampler that balances exploration and stability for RL post-training of flow-matching models via a novel posterior-mean approximation.
-
One-Forcing: Towards Stable One-Step Autoregressive Video Generation
One-Forcing augments DMD with a GAN loss to enable stable one-step causal autoregressive video generation, reporting a VBench score of 83.76 as SOTA among one-step methods.
-
Bernini: Latent Semantic Planning for Video Diffusion
Bernini is a framework that uses an MLLM planner to output semantic representations for a DiT renderer to generate or edit videos, reporting SOTA benchmark performance.
-
One Sentence, One Drama: Personalized Short-Form Drama Generation via Multi-Agent Systems
A hierarchical multi-agent framework converts a single sentence into a short drama using debate-based scripting, 3D-grounded first frames for spatial consistency, and multi-stage reviewer loops.
-
EasyVFX: Frequency-Driven Decoupling for Resource-Efficient VFX Generation
EasyVFX decouples VFX generation via frequency-aware Mixture-of-Experts and test-time training to achieve realistic effects with limited resources.
-
Dynamic Video Generation: Shaping Video Generation Across Time and Space
DVG dynamically selects content-aware spatio-temporal acceleration strategies for diffusion-based video generation, delivering up to 7x speedup with near-lossless quality on models like HunyuanVideo.
-
Temporal Aware Pruning for Efficient Diffusion-based Video Generation
TAPE applies temporal-aware token pruning with smoothing, reselection, and timestep scheduling to speed up video diffusion models while preserving visual fidelity and coherence.