# Why Local LLMs Suddenly Slow Down at Long Context

Page: https://stenobird.com/podcast/tech-stories-tech-brief-by-hackernoon-6365648/why-local-llms-suddenly-slow-down-at-long-context
Text version: https://stenobird.com/podcast/tech-stories-tech-brief-by-hackernoon-6365648/why-local-llms-suddenly-slow-down-at-long-context.md
Podcast: [Tech Stories Tech Brief By HackerNoon](https://stenobird.com/podcast/tech-stories-tech-brief-by-hackernoon-6365648)
Published: 2026-06-28T16:00:56+00:00
Episode link: https://share.transistor.fm/s/ecca952f
Audio file: https://media.transistor.fm/ecca952f/01bdbf67.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/tech-stories-tech-brief-by-hackernoon-6365648/episodes/why-local-llms-suddenly-slow-down-at-long-context
Duration seconds: 331

## Resource

This story was originally published on HackerNoon at: https://hackernoon.com/why-local-llms-suddenly-slow-down-at-long-context . Your local LLM runs fine until it doesn't. A look at KV cache spilling from VRAM into shared memory, and why it happens silently on Windows. Check more stories related to tech-stories at: https://hackernoon.com/c/tech-stories . You can also check exclusive content about #local-llms , #llama.cpp , #kv-cache , #vram , #gpu , #local-inference , #machine-learning , #hackernoon-top-story , and more. This story was written by: @speederx . Learn more about this writer by checking @speederx's about page, and for more stories, please visit hackernoon.com . Your local LLM runs fine until the context fills up past a certain point - then generation speed can drop by ~50%. The cause is the KV cache spilling out of VRAM into slower shared memory. On Windows it happens silently, with no out-of-memory error to warn you.

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/tech-stories-tech-brief-by-hackernoon-6365648/episodes/why-local-llms-suddenly-slow-down-at-long-context/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/tech-stories-tech-brief-by-hackernoon-6365648/why-local-llms-suddenly-slow-down-at-long-context.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.