Ashish Rawat (@eashish93) / X

Ashish Rawat

2,684 posts

Ashish Rawat

@eashish93

✍️ kitful.ai (ai blogging) 🧹 signupscore.com

Joined March 2015

Ashish Rawat
@eashish93
5h
GPT 5.6 Sol is really a slop model, writing too much code. After I refactored the codebase with Opus 5 it nearly removed 50% of the code from that. Going from 28 sql tables to just 12 tables and the code and tests against it all removed by Opus 5. The only good thing about
Ashish Rawat
@eashish93
9h
They are literally trying hard to replace frontier models. 😀
Sakana AI
@SakanaAILabs
9h
Announcing the Claude Code-compatible interface for our new Fugu-Ultra v1.1! 🐡 Put a dynamically coordinated team of frontier models to work inside the coding workflow you already know. Instead of relying on a single model to write, debug, and execute your code, you can now
Ashish Rawat
@eashish93
10h
Opus 5 really doesn’t work correctly on first call. But it’s really a good model for planning and present your ideas. I think it’s definitely quantized from fable.
Ashish Rawat
@eashish93
Jul 25
That’s more like 100 shots
Alex Ermolov
@alex_erm
Jul 24
Opus 5, snowboarder test, one shot. On par with Fable, ahead of every other model I've run. No visual defects on the first pass, and the sliding physics feel right.
00:00
Ashish Rawat
@eashish93
Jul 25
Opus 5 says no.
Claude
@claudeai
Jul 24
Replying to @claudeai
On several coding and knowledge work evaluations, Opus 5 is the new state-of-the-art:

New to X?