# Comparisons to previous Gen "Frontier"? — retained record

> Retained reddit record from reddit:LocalLLaMA.

Canonical: https://highsignal.app/data/records/9b041c4658c0d6cc

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:LocalLLaMA)  published 2026-11-10 18:48 ingested 2026-11-11 02:42

[original source ↗](https://www.reddit.com/r/LocalLLaMA/comments/1xsv4r7/comparisons_to_previous_gen_frontier/)

## retained text



```
There are a lot of comparisons of open-weight models to the *current* frontier, especially comparisons to Fable/Opus 5.5. I have been sticking with Opus 4.8 as I feel it behaves in a way that I appreciate much more than the 5.x Gen. It asks a lot more questions and doesn't assume nearly as much.

  So, my question is, are there comparison indexes of recent open-weight models against Opus 4.8? It appears that Opus 4.8 is generally ignored now on artificialanalysis.ai (as is Qwen3.8-Flash-Next).
```



document fetched 2026-11-11 00:42 · canonical https://reddit.com/r/LocalLLaMA/comments/1xsv4r7/comparisons_to_previous_gen_frontier
