Comparing Fable and 10 other LLMs on refactoring a LangGraph god node

https://wtf.korridzy.com/twilight-of-the-gods/

by Korridzy • 20 days ago

47 20 20 days ago

wtf.korridzy.com

azalemeth• 20 days ago

For what it is worth, your site has been blocked by the UK's national cyber security centre on my current connection:

    This site may be associated with malicious activity or malware.
     Access to this site has been blocked by the Protective DNS Service
    Site: wtf.korridzy.com
    Please contact your local Network Administrator or IT support if you require further assistance


    Look up this site on Cloudflare Radar [1] for more information.

I would wear this as a badge of pride!

[1] https://radar.cloudflare.com/domains/domain/wtf.korridzy.com

graemep• 20 days ago

It looks like its blocked by Cloudflare Radar.

I do not think the NCSC blocks sites. I have never come across a site block ordered by them. Most blocks I see are by sites themselves blocking Uk visitors for GDPR or OSA reasons.

yorwba• 20 days ago

They offer a service to let them block sites for you if you want: https://www.ncsc.gov.uk/information/pdns

Korridzy3• 20 days ago

I can see 7 visitors from the UK in the last 30 minutes. It looks like no restrictions.

rc1• 20 days ago

It’s an opt in service. Not the UK as a whole blocked.

llm_nerd• 20 days ago

It is blocked by Cloudflare, and it looks like the reason is simply that it's a super new domain so it's just an overly cautious thing.

Korridzy3• 20 days ago

A pity. I've really made this blog a couple days ago. This is my first post in it :)

nijave• 20 days ago

Any chance you'll retest with latest iterations of models?

Opus 4.8, GLM-5.2, Kimi K2.7, etc

Korridzy3• 20 days ago

It’s quite possible I’ll do that with a new work-related task. I regularly run similar experiments for myself, so I decided to write about one of them.

holoduke• 20 days ago

Even for a ui change fable complained about policy issues. This model is unusable

overgard2• 20 days ago

I had a pretty rough experience with Fable today. At first, Claude showed a message that Fable was available again, but if I typed /model it showed the same "disabled" state for the export controls, so that was sloppy. After I got it working (I had to login again), I asked it to do a security audit of a server I've written, and it spent a few tokens and then "flagged" the request and downgraded to Opus 4.8. Opus 4.8 just kind of glazed me for writing secure code without finding anything important.

By the way, I ran ccusage afterwards and if I had been paying API rates I would have apparently spent $6.88 for that.

Amazing stuff Anthropic.

clickclas• 20 days ago

https://www.anthropic.com/news/redeploying-fable-5

The US government has forced them to make the filters even stricter than they already were. It feels meaningfully worse than during its initial release.

Amazing stuff, Trump voters.

And amazing stuff to Amazon's CEO, who threw a tantrum on the basis of misinformation and created this whole mess and has how resulting in permanent government handicapping of American models. Genius move.

overgard2• 20 days ago

I don't like any of those people so I'm not disagreeing, but I feel like this is 100% on anthropic for their constant doom trolling. They're not trying to be "responsible", they just want to scare people and it worked.

whythismatters• 20 days ago

Same story here. Wanted to give Fable a spin, session got flagged mid-execution of the plan (some innocent react-native code), things fell back to Opus, and then my "current session" usage limit ran out. So I ended up with a partial and broken Fopus implementation.

andersmurphy• 20 days ago

Wait what? gpt5.5 is better than fable. I thought fable was the endtimes?!

Wonder if this is because fable swirches down to opus for somethings without telling you?

reckless• 20 days ago

Takeaways Which model to use for generating architecture.

The simpler the decision you need to make, the more readily you can just take a proposal from Fable or GPT-5.5. And now that Fable is unavailable outside the US, the choice between Opus and GPT is far from obvious. As a quick default I'd lean toward GPT.

sroerick• 20 days ago

I'm curious how folks here are using LangGraph

Comparing Fable and 10 other LLMs on refactoring a LangGraph god node

Related Stories