Why Is Video Still So Expensive? A Survey of Inference-Efficiency Mechanisms in Video and Audiovisual LLMs

arXiv:2609.10355 · cs.CV, cs.CL, cs.MM · Submitted 2026-09-09 · Read on arXiv

cs.CV, cs.CL, cs.MM

Submitted: 2026-09-09

Updated: 2026-09-09

Comments: Supplementary material at https://www.killian-steunou.com/videollm-survey/static/pdfs/videollm_survey_supplementary.pdf

Code: https://github.com/momentslab/awesome-efficient-videollm

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Accuracy — 1× in this paper · explained in 1 episode
• Grounding — 1× in this paper · explained in 3 episodes

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models