In the news
Why Is Video Still So Expensive? A Survey of Inference-Efficiency Mechanisms in Video and Audiovisual LLMs
arXiv cs.AI · Published · 3 min read
In 30 seconds
- What happened
- Survey identifies why video language models remain computationally expensive and catalogs efficiency mechanisms across frame sampling, encoding, and token reduction stages.
- Why it matters
- Engineers deploying video LLMs in real-time, mobile, or resource-constrained environments need to understand cost-reduction tradeoffs.
- Watch out
- Survey covers heterogeneous cross-paper evidence without standardized evaluation; audiovisual efficiency gaps remain largely unaddressed.
- llm
- language model
- retrieval
- prompt
- inference
The patterns behind this
Each one covers how the technique works, when it earns its cost, and where it breaks.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.