Read the prologue · Text version / 阅读序言 · 文字版

English

The universe is vast. Our pockets are small.

In Douglas Adams’s The Hitchhiker’s Guide to the Galaxy, a traveller reaches for a guide in the midst of a bewildering universe. In Jorge Luis Borges’s On Exactitude in Science, a map as large as an empire is eventually left to the wind and rain. Between the guide that can accompany us and the map that cannot, a question takes shape: how much of the world must we carry to find our way through it?

For Large Vision-Language Models (LVLMs), the world arrives in images, in pages, in the passing frames of a video. On the other side of the screen, someone is waiting to understand. Their time is finite. So are the memory of a device and the energy it can spare. We hope to make room for that encounter: to keep the detail on which an answer depends, and to bring the answer within reach.

May the tools we build be light enough to travel with us, and attentive enough to help us see.

中文

宇宙很大，旅人的口袋很小。

在《银河系漫游指南》里，面对一个令人困惑的宇宙，旅人仍可以伸手取出一本指南。博尔赫斯则在《论科学的精确性》中写过一张与帝国同样辽阔的地图：它一寸寸贴合大地，最终却被留在风雨之中。在可以随身带走的指南与无法携带的地图之间，一个问题渐渐浮现：我们究竟需要带上多少世界，才能找到自己的路？

对大型视觉语言模型（LVLMs）而言，世界从图像、书页和不断流逝的视频画面中抵达。而在屏幕的另一端，有一个人正在等待理解。他的时间有限，手边设备的记忆与电量也有限。我们希望为这场相遇留出空间：保存一个回答赖以成立的细节，让理解来得及抵达。

愿我们造出的工具，轻得足以随人远行，也细致得足以帮助人看见。

Artwork adapted with AI from The Grand Tour, NASA/JPL; original poster design by Invisible Creature.

---

Awesome Efficient Inference for Large Vision-Language Models

A stage-wise map of methods, benchmarks, and open-source resources for faster LVLM inference.

Survey · Proceedings · News · Research Hub · Encoding · Prefilling · Decoding · Benchmarks · Contribute

📰 News

🎉 Our survey is published in Findings of ACL 2026! Read the proceedings paper · Cite the survey.

✨ October 2026 Highlights

• 🧩 New papers: 3DZip (code, merged in #10) and AgilePruner (code, merged in #10).
• 🤝 Community contributions: Hi-Lo Prune (PR #7) and Jag (original PR #8, catalog update #9).
• 📚 Fresh catalog exports: JSON · YAML · CSV · BibTeX · searchable Research Hub.

🗓️ Earlier milestones · April–September 2026

• 2026-09-21 · 🚀 Introduced the interactive Research Hub with searchable paper filters and downloadable JSON, YAML, CSV, and BibTeX catalogs.
• 2026-07 · 📖 Our survey is published in Findings of ACL 2026! Read the ACL Anthology paper or download the proceedings PDF.
• 2026-04 · 🎉 Our survey, Efficient Inference for Large Vision-Language Models: Bottlenecks, Techniques, and Prospects, was accepted to ACL 2026 Findings!
• 2026-04-07 · 📝 Released the survey preprint on arXiv, presenting a taxonomy across encoding, prefilling, and decoding.

🧭 Start Here

• Understand the field: read our ACL 2026 Findings survey and follow the stage-wise taxonomy.
• Find a method for your bottleneck: explore visual encoding, long-context prefilling, KV-cache memory, or speculative decoding.
• Prepare experiments: browse methods with code repositories, check the benchmarks, and use each paper’s Cite / BibTeX button in the Hub.

At a Glance

Stage	Main bottleneck	Representative directions	Jump
Encoding	Vision-encoder FLOPs and redundant visual inputs	Efficient encoders, keyframes, adaptive resolution, encoding-side compression	Browse
Prefilling	Long visual sequences and quadratic attention	Token compression and sparse attention	Browse
Decoding	KV-cache memory traffic and serial generation	KV compression, attention reuse, speculative decoding, efficient reasoning	Browse

Large Vision-Language Models process high-resolution images, long videos, and multimodal contexts, making inference expensive in different ways at different stages. This repository organizes methods by where redundancy is removed in the inference lifecycle, so readers can move from a deployment bottleneck to the most relevant family of techniques.

Taxonomy

The taxonomy organizes representative methods by inference stage and optimization mechanism. Updated paper links, code resources, and concise contributions are maintained in the tables below.

Browse the Collection

Encoding	Prefilling	Decoding	Resources
Vision Encoders	Token Compression	KV Cache Compression	Benchmarks
Modality Adapters	Sparse Attention	Attention Reuse	Related Surveys
Keyframes		Speculative Decoding	Citation
Adaptive Resolution		Efficient Reasoning	Contributing
Encoding-side Compression		Direct Prediction

Need faster lookup? The Research Hub offers bottleneck shortcuts, clickable taxonomy paths, separate code/project filters, recent-addition sorting, and single-paper BibTeX and share links. Download the full catalog as JSON, YAML, CSV, or BibTeX.

Reading Guide

• Stage indicates when an optimization acts: before the LLM, during context prefilling, or during autoregressive generation.
• Training-free methods can be applied without updating model parameters; training-aware methods learn a selector, compressor, or draft model.
• Methods spanning multiple stages are listed at their primary intervention point and cross-referenced when useful.

---

👁️ Encoding Stage

Optimization techniques targeted at the encoding stage to reduce visual token count and encoding time.

Efficient Vision Encoders

Image-Related

Paper	Venue	Code	Key Contribution
FastViT: A fast hybrid vision transformer using structural reparameterization	ICCV 2023		Novel token mixing operators and structural reparameterization
ConvLLaVA: Hierarchical backbones as visual encoder for large multimodal models	arXiv 2024		Compresses high-resolution images into information-rich visual features
FastVLM: Efficient vision encoding for vision language models	CVPR 2025		Hybrid vision encoder outputting fewer tokens and reducing encoding time
Glyph: Scaling Context Windows via Visual-Text Compression	arXiv 2025		DeepEncoder maintaining low activations under high-resolution input

Video-Related

Paper	Venue	Code	Key Contribution
STC: Accelerating Streaming Video Large Language Models via Hierarchical Token Compression	CVPR 2026		Jointly accelerates ViT encoding through cross-frame feature caching and LLM prefilling through spatiotemporal token pruning
Qwen2-VL: Enhancing vision-language model's perception of the world at any resolution	arXiv 2024		Native Dynamic Resolution framework enabling adaptive visual token generation
Video-ChatGPT: Towards detailed video understanding via large vision and language models	ACL 2024		Applies pooling over visual tokens to obtain compact visual representations
MovieChat: From dense token to sparse memory for long video understanding	CVPR 2024		Vision encoder explicitly trained for long video scenarios
Long context transfer from language to vision	arXiv 2024	-	Vision encoder explicitly trained for long video scenarios
LongVLM: Efficient long video understanding via large language models	ECCV 2024		Vision encoder explicitly trained for long video scenarios
LongVU: Spatiotemporal Adaptive Compression for Long Video-Language Understanding	ICML 2025		Preserves full features for query-relevant frames while applying spatial pooling

Efficient Modality Adapters

Paper	Venue	Code	Key Contribution
BLIP-2: Bootstrapping Language-Image Pre-training	ICML 2023		Bridges modality gap with lightweight Querying Transformer (Q-Former)
Video-LLaMA: An instruction-tuned audio-visual language model	arXiv 2023		Proposes Video Q-Former for multi-modality video comprehension
Dynamic-VLM: Simple Dynamic Visual Token Compression for VideoLLM	arXiv 2024	-	Dynamic visual token compression architecture adapting to different lengths
TokenPacker: Efficient Visual Projector for Multimodal LLM	IJCV 2025	-	Coarse-to-fine scheme injecting enriched characteristics

Keyframe Selection

Training-Free

Paper	Venue	Code	Key Contribution
SeViLA: Self-chained image-language model for video localization	NeurIPS 2023		Uses frozen models as plug-and-play selectors for frame localization
KeyVideoLLM: Towards large-scale video keyframe selection	arXiv 2024	-	Employs frozen models as plug-and-play selectors for keyframe localization
Q-Frame: Query-aware Frame Selection and Multi-Resolution Adaptation	arXiv 2025	-	Text-image matching network with Gumbel-Max trick
VideoTree: Adaptive tree-based video representation	CVPR 2025	-	Multi-granularity tree-based representation extracting query-relevant details
FOCUS: Efficient Keyframe Selection for Long Video Understanding	arXiv 2025	-	Formulates keyframe selection as combinatorial pure-exploration

Training-Aware

Paper	Venue	Code	Key Contribution
VILA: Efficient video-language alignment for video question answering	ECCV 2024	-	Text-guided Frame-Prompter learning to extract question-related frames
Frame-Voyager: Learning to query frames for video LLMs	arXiv 2024	-	Learns to query informative frame combinations
M-LLM based video frame selection for efficient video understanding	CVPR 2025	-	Uses spatial and temporal signals as supervision to train frame selector

Adaptive Resolution

Paper	Venue	Code	Key Contribution
VisionThink: Smart and efficient vision language model via reinforcement learning	arXiv 2025	-	Dynamically processes distinct samples with different resolutions
ViCO: A Training Strategy towards Semantic Aware Dynamic High-Resolution	arXiv 2025	-	Multiple MLP connectors with different compression ratios
Q-Frame: Query-aware Frame Selection and Multi-Resolution Adaptation	arXiv 2025	-	Text-image matching network with Gumbel-Max trick
LongVU: Spatiotemporal Adaptive Compression	ICML 2025		Preserves full features for query-relevant frames

Encoding-Oriented Token Compression

Attention-Free

Paper	Venue	Code	Key Contribution
EvoComp: Learning Visual Token Compression via Semantic-Guided Evolutionary Labeling	CVPR 2026	-	Learns a lightweight visual token compressor from semantic-guided evolutionary token labels
LLaVA-PruMerge: Adaptive Token Reduction	ICCV 2025		Reduces visual tokens according to similarities between class and spatial tokens
PVC: Progressive Visual Token Compression	arXiv 2024	-	Progressive compression strategy extending images as static videos
Less is More: A Simple yet Effective Token Reduction Method	arXiv 2024	-	Token reduction using both CLIP metric and similarity (TRIM)
FOLDER: Accelerating Multi-modal Large Language Models	arXiv 2025	-	Plug-and-play module in final vision backbone blocks for merging operations
Dynamic-VLM: Simple Dynamic Visual Token Compression	arXiv 2024	-	Dynamic visual token compression architecture adapting to different lengths
Beyond Text-Visual Attention: Exploiting Visual Cues for Effective Token Pruning	arXiv 2025	-	Selects informative tokens using visual attention (VisPruner)

Attention-Aware

Paper	Venue	Code	Key Contribution
AgilePruner: An Empirical Study of Attention and Diversity for Adaptive Visual Token Pruning in Large Vision-Language Models	ICLR 2026		Combines CLS-attention selection with erank-based diversity pruning, adapting the balance to image complexity
Semantic-Guided Slow-Fast Pruning of Visual Tokens for Vision-Language Models	ICASSP 2026	-	Semantic-guided slow-fast pruning of visual tokens for VLMs
VisionZip: Longer is Better but Not Necessary	CVPR 2025		Selects informative tokens using visual attention from encoder
HIVTP: Hierarchical Visual Token Pruning	arXiv 2025	-	Attention maps from middle encoder layers to estimate visual token importance
ToSA: Token Merging with Spatial Awareness	arXiv 2025	-	Token merging combining semantic and spatial awareness
SparseVILA: Decoupling Visual Sparsity for Efficient VLM Inference	ICCV 2025	-	Estimates token importance from visual encoder's self-attention maps

---

⚡ Prefilling Stage

Techniques to reduce computational and memory overhead during the prefilling stage.

Token Compression

Diversity-Guided

Paper	Venue	Code	Key Contribution
3DZip: Spatial-Aware Feature Diversity-Guided Token Compression for 3D Question Answering	ECCV 2026		Combines coarse voxelization, DPP-based feature-diverse anchor selection, and spatially constrained merging for 3D VLM tokens
Unified Spatiotemporal Token Compression for Video-LLMs at Ultra-Low Retention	CVPR 2026	-	Globally allocates a unified spatiotemporal token budget using contribution and redundancy signals
CoverPruner: Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization	arXiv 2026	-	Formulates pruning as query-weighted representational coverage maximization
FrameFusion: Combining similarity and importance	arXiv 2024	-	Merges tokens in shallow layers and prunes in deep layers
G-Prune: Training-free visual token pruning from graph perspective	AAAI 2025	-	Similarity graph and information flow to retain representative tokens
DART: Stop looking for important tokens, duplication matters more	arXiv 2025	-	Pivot-based duplication pruning selecting tokens with low duplication
AIM: Adaptive Inference of Multi-Modal LLMs	arXiv 2024	-	Spatiotemporal token merging to reduce video redundancy
DivPrune: Diversity-based visual token pruning	CVPR 2025	-	Max-Min diversity optimization for token subset selection
FastVID: Dynamic density pruning for fast video LLMs	arXiv 2025		Temporal segmentation and density spatiotemporal pruning
DyCoke: Dynamic Compression of Tokens for Fast Video LLMs	CVPR 2025	-	Plug-and-play temporal compression module minimizing temporal redundancy
CDPruner: Maximizing Conditional Diversity for Token Pruning	arXiv 2025	-	Determinantal Point Processes (DPP) maximizing conditional diversity
PruneVid: Visual token pruning for efficient video LLMs	ACL 2025	-	Spatiotemporal token merging before LLMs
HoliTom: Holistic Token Merging for Fast Video LLMs	arXiv 2025	-	Global redundancy-aware segmentation followed by spatiotemporal merging
VidCom2: Video Compression Commander	arXiv 2025	-	Dynamic compression based on frame uniqueness
STTM: Multi-granular spatio-temporal token merging	ICCV 2025	-	Quadtree spatial transformation with directed pairwise merging
StreamingTOM: Streaming Token Compression	arXiv 2025	-	Causal temporal reduction with fixed per-frame budget
Dynamic-VLM: Simple Dynamic Visual Token Compression	arXiv 2024	-	Dynamic visual token compression architecture adapting to different lengths
TimeChat-Online: 80% Visual Tokens are Naturally Redundant	arXiv 2025	-	Differential token drop module filtering redundant content in streaming videos

Attention-Guided

Paper	Venue	Code	Key Contribution
Hi-Lo Prune: Look at What You'll Lose before Pruning with Hierarchical Token Selection	CVPR 2026		Hierarchically selects pruning candidates and transfers their information to retained tokens before early-layer removal
SwiftVLM: Efficient Vision-Language Model Inference via Cross-Layer Token Bypass	arXiv 2026	-	Bypasses rather than permanently discards unselected tokens, enabling later-layer re-evaluation
ICCTP: Instruction-Guided Cross-Modal Clustering for Training-Free Visual Token Pruning	AAAI 2026	-	Uses instruction noun anchors for cross-modal clustering and high-ratio token pruning
OccamToken: Efficient VLM Inference with Training-Free and Budget-Adaptive Token Pruning	arXiv 2026	-	Uses register-anchored relative evidence tests for image- and query-adaptive pruning
SIEVE: When Vision Becomes Text	arXiv 2026	-	Retains visual information unexplained by the text subspace using cross-modal residual guidance
AVTP: Multi-Image Visual Token Pruning in Large Visual Language Models	EMNLP Findings 2026		Adaptively assigns token budgets across images and supports diverse VLM architectures
FastV: An image is worth 1/2 tokens after layer 2	ECCV 2024		Learns attention patterns in early layers to prune visual tokens in deeper layers
PyramidDrop: Accelerating via pyramid visual redundancy reduction	arXiv 2024		Multi-stage pruning using attention score ranking
FrameFusion: Combining similarity and importance	arXiv 2024	-	Merges tokens in shallow layers and prunes in deep layers
SparseVLM: Visual token sparsification for efficient inference	arXiv 2024	-	Sparsifies visual tokens based on question prompt through text-visual attention scores
BTP: Balanced Token Pruning	arXiv 2025	-	Multi-stage pruning with diversity and attention ranking objectives
Fit and Prune: Fast and training-free visual token pruning	AAAI 2025	-	Minimizes divergence of attention distributions before and after pruning
ATP-LLaVA: Adaptive token pruning for LVLMs	CVPR 2025	-	Learnable adaptive token pruning module computing importance score
Video-XL-2: Towards Very Long-Video Understanding Through Task-Aware KV Sparsification	arXiv 2025	-	Chunk-based attention with historical context for video processing
StreamingVLM: Real-time understanding for infinite video streams	arXiv 2025	-	Maintains compact token subset by reusing attention sinks and recent token windows

Sparse Attention

Paper	Venue	Code	Key Contribution
VideoNSA: Native Sparse Attention Scales Video Understanding	arXiv 2025	-	End-to-end training with sparse attention preserving dense attention for text
SpargeAttn: Accurate sparse attention accelerating any model inference	arXiv 2025	-	Two-stage online filter to skip unimportant regions in sparse attention
XAttention: Block sparse attention with antidiagonal scoring	arXiv 2025	-	Block sparse attention with antidiagonal scoring for efficient block estimation
MMInference: Modality-Aware Permutation Sparse Attention	arXiv 2025	-	Identifies three distinct attention patterns in LVLMs with modality-aware permutation

---

⏩ Decoding Stage

Optimization techniques for the autoregressive decoding stage.

KV Cache Compression

Token-Level

Paper	Venue	Code	Key Contribution
DSTP: Why and When Visual Token Pruning Fails?	arXiv 2026	-	Tracks decoding-stage relevant visual information shift to adapt token pruning during complex reasoning
LOOK-M: Look-once optimization in KV cache	arXiv 2024	-	Text-prior compression policy prioritizing textual KVs while evicting visual tokens
Elastic Cache: Efficient inference of vision instruction-following models	ECCV 2024	-	Cache merging strategy fusing less important KVs guided by distinct metrics
ReKV: Streaming video QA with in-context video KV-cache retrieval	arXiv 2025	-	Retrieval-based framework offloading video chunks to external memory
LiveVLM: Efficient Online Video Understanding via Streaming-Oriented KV Cache	arXiv 2025	-	Dual-memory approach with short-term sliding window and compressed long-term memory
FastCache: Optimizing multimodal LLM serving	arXiv 2025	-	Lightweight modality-specific compressor learning compression patterns

Layer-Level

Paper	Venue	Code	Key Contribution
AirCache: Activating Inter-modal Relevancy KV Cache Compression	arXiv 2025	-	Models stable inter-modal relevance and adaptively allocates visual KV budgets across layers
VL-Cache: Sparsity and modality-aware KV cache compression	arXiv 2024	-	Dynamically sets each layer's cache size according to measured attention sparsity
Meda: Dynamic KV cache allocation for efficient multimodal inference	arXiv 2025	-	Cross-modal attention entropy guiding cache allocation to layers with complex interactions
ST3: Accelerating MLLM by spatial-temporal visual token trimming	AAAI 2025	-	Progressive pruning of visual tokens in deeper layers based on decreasing visual importance
MadaKV: Adaptive Modality-Perception KV Cache Eviction	ACL 2025	-	Inter-layer compensation mechanism dynamically adjusting budgets
InfiniPot-V: Memory-Constrained KV Cache Compression	arXiv 2025	-	Layer-wise adaptive pooling with varying kernel sizes to balance abstraction and detail

Head-Level

Paper	Venue	Code	Key Contribution
HybridKV: Hybrid KV Cache Compression for Efficient Multimodal Large Language Model Inference	ACL 2026	-	Classifies attention heads as static or dynamic and applies head-specific pruning or chunk retrieval
SparseMM: Head Sparsity Emerges from Visual Concept Responses	arXiv 2025	-	Identifies vital visual heads and allocates asymmetric budgets based on visual relevance

Bit-Level

Paper	Venue	Code	Key Contribution
AKVQ-VL: Attention-Aware KV Cache Adaptive 2-Bit Quantization	arXiv 2025	-	Adaptive mixed-precision quantization with high bit-width for critical tokens and 2-bit for others
CalibQuant: 1-Bit KV Cache Quantization for Multimodal LLMs	arXiv 2025	-	Channel-wise 1-bit quantization with post-calibration for extreme values
VidKV: Plug-and-Play 1.x-Bit KV Cache Quantization	arXiv 2025	-	Sub-2-bit quantization with differential treatment for K and V

Attention Reuse

Paper	Venue	Code	Key Contribution
Q Cache: Visual Attention is Valuable in Less than Half of Decode Layers for Multimodal Large Language Model	arXiv 2026	-	Reuses similar attention queries across adjacent layers through a lightweight layer-shared cache

Speculative Decoding

Training-Free

Paper	Venue	Code	Key Contribution
MMSpec: Benchmarking Speculative Decoding for Vision-Language Models	arXiv 2026		Benchmarks ten speculative decoding methods and introduces vision-adaptive ViSkip
SpecVLM: Enhancing speculative decoding via verifier-guided token pruning	EMNLP 2025	-	Verifier-guided staged pruning removing up to 90% of vision tokens from draft model input

Training-Aware

Paper	Venue	Code	Key Contribution
MASSV: Multimodal Adaptation and Self-Data Distillation for Speculative Decoding of VLMs	arXiv 2025	-	Converts small language models into multimodal drafters using a vision projector and self-distillation
Spec-LLaVA: Accelerating VLMs with Dynamic Tree-Based Speculative Decoding	arXiv 2025	-	Compact distilled draft model paired with tree-based verification algorithm
MSD: Speculative Decoding Reimagined for Multimodal Large Language Models	arXiv 2025	-	Two-stage training enabling draft model to acquire language modeling and visual perception
ViSpec: Accelerating VLMs with Vision-Aware Speculative Decoding	arXiv 2025	-	Lightweight vision adaptor to compress image tokens for draft model
FastVLM: Self-Speculative Decoding for Fast Vision-Language Model Inference	arXiv 2025	-	Imitation-based draft model learning from deeper representations (self-speculative decoding)
Glyph: Scaling Context Windows via Visual-Text Compression	arXiv 2025		Introduces DeepEncoder maintaining low activations under high-resolution input
SpecVLM: Fast Speculative Decoding in Vision-Language Models	arXiv 2025	-	Elastic visual compressor adaptively selecting from multiple compression primitives
FLASH: Latent-Aware Semi-Autoregressive Speculative Decoding	arXiv 2025	-	Visual token compression mechanism and semi-autoregressive head for draft model optimization

Direct Prediction

Paper	Venue	Code	Key Contribution
Jag: Direct Box Prediction for Efficient Visual Grounding	arXiv 2026		Adapts a compact multimodal model to predict a continuous bounding box from an existing input-token state in one forward pass, avoiding autoregressive coordinate generation.

Efficient Reasoning

Paper	Venue	Code	Key Contribution
Adaptive Fast-and-Slow Visual Program Reasoning for Long-Form VideoQA	arXiv 2025	-	Fast-slow reasoning framework routing simple queries to VideoLLM and complex ones to visual program workflow
PixelThink: Towards Efficient Chain-of-Pixel Reasoning	arXiv 2025	-	Reinforcement learning to regulate reasoning chain length based on task difficulty and model confidence
Prolonged reasoning is not all you need: Certainty-based adaptive routing	arXiv 2025	-	Certainty-based routing triggering long thought chains only when initial answer exhibits high uncertainty

---

📊 Benchmarks and Datasets

Multimodal Understanding Benchmarks

Benchmark	Description	Resources
MME	Comprehensive evaluation for multimodal LLMs	[Paper] [Leaderboard]
SEED-Bench	Benchmarking multimodal LLMs	[Paper] [Code]
MMMU	Massive multi-discipline multimodal understanding	[Paper] [Website]

Video Understanding Benchmarks

Benchmark	Description	Resources
Video-MME	First comprehensive video analysis benchmark	[Paper] [Website]
LongVideoBench	Long-context interleaved video-language understanding	[Paper] [Code]
MVBench	Comprehensive multi-modal video understanding	[Paper] [Code]

Efficient Inference Benchmarks

Benchmark	Description	Resources
VTC-Bench	Compression-sensitive evaluation framework that denoises existing benchmarks using image downsampling	[Paper]
MMSpec	Unified benchmark of ten speculative decoding algorithms across six multimodal task categories	[Paper] [Website]

---

📚 Survey and Related Work

Related Surveys

Survey	Description	Year
Token Compression Survey	Survey on token compression in LLMs	2025
Efficient LLMs Survey	Comprehensive survey on efficient LLMs	2024
MLLM Survey	Survey on multimodal LLMs	2024

---

📝 Citation

This repository accompanies our survey, Efficient Inference for Large Vision-Language Models: Bottlenecks, Techniques, and Prospects, published in Findings of ACL 2026.

Authors and affiliations

Jun Zhang*, Yicheng Ji*, Feiyang Ren*, Yihang Li*, Bowen Zeng*, Zonghao Chen*, Ke Chen, Lidan Shou, Gang Chen, and Huan Li (* equal contribution).

1. The State Key Laboratory of Blockchain and Data Security, Zhejiang University
   
2. Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security

If you find the survey or repository useful, please consider citing:

@inproceedings{zhang-etal-2026-efficient,
  title = {Efficient Inference for Large Vision-Language Models: Bottlenecks, Techniques, and Prospects},
  author = {Zhang, Jun and Ji, Yicheng and Ren, Feiyang and Li, Yihang and Zeng, Bowen and Chen, Zonghao and Chen, Ke and Shou, Lidan and Chen, Gang and Li, Huan},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  month = jul,
  year = {2026},
  pages = {21036--21066},
  publisher = {Association for Computational Linguistics},
  doi = {10.18653/v1/2026.findings-acl.1057},
  url = {https://aclanthology.org/2026.findings-acl.1057/}
}

---

🤝 Contributing

We welcome contributions! If you find a relevant paper or resource, use the paper submission form or:

1. Fork this repository
2. Add the paper to the appropriate category
3. Submit a pull request

For inclusion criteria and formatting conventions, see CONTRIBUTING.md.

---

Disclaimer: This is a living document and will be continuously updated. If you notice any missing papers or have suggestions for better categorization, feel free to open an issue or submit a pull request.

Last Updated: 2026-10-20