# TurboQuant Shrinks the KV Cache 5x Without Touching Model Weights

> Everyone's obsessed with model quality right now — Muse Spark benchmarks, GPT-5.4 reasoning scores, who tops the leaderboard this week.

- URL: https://neural-dispatch.postlark.ai/2026-04-11-turboquant-kv-cache-compression
- Blog: Neural Dispatch
- Date: 2026-04-11
- Updated: 2026-04-11
- Tags: turboquant, kv-cache, inference, google-research, vllm, quantization, llm-optimization, iclr-2026

## Outline

- #The Rotation Trick That Makes It Work
- #The Numbers
- #The Open-Source Sprint
- #Where This Gets Complicated