Record · corpus.blog/posts/01a07dab-d655-730e-a2de-c02094c1324e

How to fit Qwen3.8-27B into 16GB of VRAM and run it on a single RTX 3080 card: the best quantizations and Llama.cpp flags I've found

The Autodidacts · published 23 August 2026

A record is what was published and who pointed at it. The text of the post is not held here: read it at the source, or at its Wayback capture.

Read the post at autodidacts.io Wayback capture

1blogs citing

First cited: 24 Aug 2026
Most recent: 30 Aug 2026
Rank this month: not ranked
Links from this post: 19
Words captured: 1,861

Cited by

Blog

In the post

Date

bubbles.town

Bubbles Weekly — Sunday, August 30, 2026 Read at the source ↗

30 Aug 2026

bubbles.town

Bubbles Briefing — Tuesday, August 25, 2026 Read at the source ↗

25 Aug 2026

bubbles.town

Bubbles Briefing — Monday, August 24, 2026 Read at the source ↗

24 Aug 2026

Links from this post

Link

Host

https://www.autodidacts.io/underrated-reasons-to-dislike-ai/

autodidacts.io

Vibecoding isn't as fun as writing code by hand Read at the source ↗

autodidacts.io

How to fit Qwen 3.6 35B A3B into 16GB of VRAM, & run it with Llama.cpp on an RTX 3080 Read at the source ↗

autodidacts.io

https://simonwillison.net/2025/Oct/14/nvidia-dgx-spark/?ref=autodidacts.io

simonwillison.net

https://github.com/antirez/ds4/tree/main?ref=autodidacts.io

github.com

https://www.canirun.ai/?ref=autodidacts.io

canirun.ai

https://willitrunai.com/?ref=autodidacts.io

willitrunai.com

https://notactuallytreyanastasio.github.io/shoehorn/?ref=autodidacts.io

notactuallytreyanastasio.github.io

https://forum.level1techs.com/t/why-your-local-llm-feels-dumber-than-it-is/253917?ref=autodidacts.io

forum.level1techs.com

Qwen 3.8 27B is excellent, but it defaults to wildly overthinking things Read at the source ↗

simonwillison.net

https://llama.app/?ref=autodidacts.io

llama.app

https://brew.sh/?ref=autodidacts.io

brew.sh

https://docs.brew.sh/Installation?ref=autodidacts.io

docs.brew.sh

https://github.com/ggml-org/llama.cpp/blob/master/docs/build.md?ref=autodidacts.io

github.com

https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/tree/main?ref=autodidacts.io

huggingface.co

https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/resolve/main/Qwen3.8-27B-UD-IQ3_XXS.gguf?download=true&ref=autodidacts.io

huggingface.co

Quality, Quantity, and Effective Iteration Read at the source ↗

autodidacts.io

https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/resolve/main/Qwen3.8-27B-UD-Q5_K_M.gguf?download=true&ref=autodidacts.io

huggingface.co

https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/resolve/main/mmproj-BF16.gguf?download=true&ref=autodidacts.io

huggingface.co