Record · corpus.blog/posts/01a0deab-4d56-7103-bb2c-ba9a82fdac8f
Benchmarking Qwen3.8 27B quantizations: 4-bit holds up, 1-bit collapses
quesma.com · published 26 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.
1blog citing
- First cited
- 21 Feb 2025
- Most recent
- 21 Feb 2025
- Rank this month
- not ranked
- External links
- 18
- Words captured
- 1,254
Cited by 1 blog in 1 post
One row per blog, the most recently citing first. Each shows the blog's newest citing post; open a row for the others.
Blog
Newest citing post
Date
Similar posts
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
23 Aug 2026
Running Qwen 3.8 27B on a 16 GB GPU
paulsprogrammingnotes.com
23 Aug 2026
Running the Qwen 3.8 27B Dense Model on 8GB VRAM
brainwagon.org
24 Aug 2026
Run Qwen3.8-27B-Escha-W2 on a 24GB GPU with SGLang
mindstudio.ai
25 Aug 2026
27 Aug 2026
22 Aug 2026
Running Qwen3.8-27B locally on a 12GB VRAM card
carteakey.dev
19 Aug 2026
Running Qwen3.8-27B locally on a 12GB VRAM card
carteakey.dev
19 Aug 2026
My "4-bit" quant was 6.2 bits per weight
koleslaw.ai
29 Jul 2026
24 Jul 2026
Links from this post
Link
Host
https://forum.level1techs.com/t/why-your-local-llm-feels-dumber-than-it-is/253917?ref=autodidacts.io
forum.level1techs.com
artificialanalysis.ai
github.com
modal.com
github.com
news.ycombinator.com