# Is every other model just a nerfed version of Fable? — retained record

> Retained reddit record from reddit:ClaudeCode.

Canonical: https://highsignal.app/data/records/4ff9cc16fe3463bf

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:ClaudeCode)  published 2026-10-08 23:06 ingested 2026-10-09 02:39

[original source ↗](https://www.reddit.com/r/ClaudeCode/comments/1x15lqm/is_every_other_model_just_a_nerfed_version_of/)

## retained text



```
so, anthropic recently released its models in this order: fable 5.1, then opus 5.5, sonnet 5.5, and haiku 5.5. 

i genuinely think we'll get fable 5.5 in the near future, but i'm not sure. i'm not really an ml guy, but i'm curious about how anthropic actually develops its models behind the scenes.

do they put most of their effort into one flagship model, then, once they make a breakthrough or get some impressive rl results, turn it into a sota model? and then do they nerf it down or distill it into smaller versions for different use cases, performance levels, and price points? for example, opus for expert work, sonnet for general use, and haiku for speed?

or is each model essentially its own project, with a different architecture, a separate research team, and different goals?

sorry if this is a dumb question, i'm just curious about how things work behind the scenes. maybe someone who works at anthropic or in the field could shed some light on this. btw, i asked claude about it and got hit with "I don’t have inside knowledge of how I was trained..." this has to be a joke!!!

ps: is fable just a nerfed version of mythos? haha
```



document fetched 2026-10-09 00:42 · canonical https://reddit.com/r/ClaudeCode/comments/1x15lqm/is_every_other_model_just_a_nerfed_version_of
