# Unconvinced by Claude Fable 5.2 — retained record

> Retained reddit record from reddit:ClaudeAI.

Canonical: https://highsignal.app/data/records/beb12d68bcd0d7be

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:ClaudeAI)  published 2026-11-10 18:45 ingested 2026-11-11 02:37

[original source ↗](https://www.reddit.com/r/ClaudeAI/comments/1xsv2s0/unconvinced_by_claude_fable_52/)

## retained text



```
As the title said, I have reservations regarding Fable 5.2. I have been extensivly using Opus 5.5, and for my usercase (Civil Engineering) it has a very down to earth, straight to the point attitude, very little in terms of hallucinations or false information presentation. If it doesn't know something it sais so, if the analysis takes long it takes long, and i like and respect that. 

At the same time I have been using Fable 5.2 for some 16 hours now (since the release). Looks fast on the surface, professional in the answers but it cuts a lot of corners, gives half answers and gives them with incredible convinction. The assumptions are wrong and the more I use it the more I have to go back and check every single answer it gave me twice, to the point that I almost feel I have gone back to Opus 5.0.

It is still a great model, but I have grown weary of it quickly, and am almost scared to give it more demanding task for the fear of making mistakes. 

All of my workflows just feel better with Opus 5.5 and it might be me just being used to its style. So I wanted some opinions on what do you guys feel this early in the release.?
```



document fetched 2026-11-11 00:21 · canonical https://reddit.com/r/ClaudeAI/comments/1xsv2s0/unconvinced_by_claude_fable_52
