# How Was your experience running Q1 Models on your machine? — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/535a9c1cdf356ccc

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-09-20 00:36 ingested 2026-09-21 02:38

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1wl2ey8/how_was_your_experience_running_q1_models_on_your/)

## retained text



```
Also Q2 and Q3

are they really that bad as people around say? or it depends highly on the model architecture and the quantize technique?

is there Big difference between UD-Q and IQ?

is Q1 really worst Quantization Available or no?

and who would win? Small Q4 Model? or Huge 200B+ MoE Q1 model?
```



document fetched 2026-09-21 00:37 · canonical https://reddit.com/r/LocalLLaMA/comments/1wl2ey8/how_was_your_experience_running_q1_models_on_your
