V-Rex (KV-cache compression): current frontier — recent, not yet superseded in the knowledge base. 0 paper(s) critique it, 0 beat it on benchmarks — not ranked as a superseded baseline. Sub-problem: cluster led by ReKV. Newer alternatives in the same sub-problem include MuKV, KVCapsule, Decoupled Streaming Cache (DSCache), LightKV, Hierarchical Adaptive Eviction (HAE).

KV-cache compression

V-Rex

V-Rex: Real-Time Streaming Video LLM Acceleration via Dynamic KV Cache Retrieval

On the frontierfirst seen Dec 13, 2025

Current frontier — recent, not yet superseded in the knowledge base

0 papers critique it · 0 beat it on benchmarks

What to use instead

Recent methods in the same sub-problem, not yet superseded in the knowledge base — arXiv benchmark leaders, not vetted production recommendations.

MuKVMuKV: Multi-Grained KV Cache Compression for Long Streaming Video Question-Answering
May 21, 2026
KVCapsuleKVCapsule: Efficient Sequential KV Cache Compression for Vision-Language Models with Asymmetric Redundancy
May 14, 2026
Decoupled Streaming Cache (DSCache)Decouple and Cache: KV Cache Construction for Streaming Video Understanding
May 3, 2026
LightKVMake Your LVLM KV Cache More Lightweight
May 1, 2026
Hierarchical Adaptive Eviction (HAE)Hierarchical Adaptive Eviction for KV Cache Management in Multimodal Language Models
Feb 2, 2026
V-RexV-Rex: Real-Time Streaming Video LLM Acceleration via Dynamic KV Cache Retrieval
Dec 13, 2025
StreamKVStreamKV: Streaming Video Question-Answering with Segment-based KV Cache Retrieval and Compression
Nov 10, 2025