Prajwal Prasad
2.8K posts
RULERS BECAME KILLERS- shame on you BJP
All India Trinamool Congress@AITCofficial
Staying true to his word, Shri @abhishekaitc REFUSED TO ABANDON the grieving family of Sanju Karmakar. Even after a vicious attack by @BJP4Bengal-backed miscreants, our National General Secretary chose not to turn back. Instead, he stood beside a family devastated by an unimaginable loss allegedly inflicted by BJP-backed hooligans. That is the difference between politics driven by compassion and politics driven by hatred. @SuvenduWB, today’s events have once again exposed the kind of politics you preside over - one rooted in intimidation, violence and vendetta.
English
Prajwal Prasad retweetledi
@thsottiaux I want a user dash board to check total tokens consumed.
English
Prajwal Prasad retweetledi
Karnataka: New Residence for Khanglus ??
Illwgal Khangladeshis escaped from Assam are now intruding Karnataka as labours.
Allegedly this video is from Mysore.
Why they are only targetting Non BJP ruled states only for shelter?
We need to win Karnataka too.
English
@thsottiaux @0xPaulius bro where is the update? its Thursday already!
English
Goal is so goated that it is the most underrated feature in codex.
Even claude code copied it.
English
We’ve been putting a lot of effort into making Claude Code more responsive & reliable.
Here’s an update on everything we’ve done:
English
Prajwal Prasad retweetledi
It took me like 2 months, but I've grown to love gpt-5.5.
You have to prompt entirely different and put some time into your agents[.]md. Now that I'm over the hump, I can't really use any other model for code.
English
Prajwal Prasad retweetledi
wait a minute 💀
they made a benchmark to test whether coding agents can handle real long horizon engineering work - repo understanding, multi file edits, tool use, debugging loops, test feedback, and keeping the system coherent across the whole task
and GPT 5.5 is already at 70% LMAO
(OpenAI already has an internally better model btw)
if this isn’t a sign of acceleration, I don’t know what is
Serena Ge (Datacurve)@serenaa_ge
Today we’re releasing DeepSWE, a new standard for agentic coding benchmarks. On public leaderboards, top models often look relatively close in capability. DeepSWE shows where they actually diverge, reflecting the realistic experience of developers in their day-to-day work.
English
Prajwal Prasad retweetledi
This is the first code bench that actually aligns with how it feels to use these models coding.
Serena Ge (Datacurve)@serenaa_ge
Today we’re releasing DeepSWE, a new standard for agentic coding benchmarks. On public leaderboards, top models often look relatively close in capability. DeepSWE shows where they actually diverge, reflecting the realistic experience of developers in their day-to-day work.
English
@orcdev share your honest feedback, comparing with codex and claude code
English
ok I'm back to Cursor
let's see what Composer 2.5 can really do
OrcDev@orcdev
is Composer 2.5 really that good? 😅 currently paying for: $100 Claude $200 Codex (open source program thank goodness ⚔️) $300 Grok Build (testing) and now I have to give Cursor another $20 too
English
English
@jturntdev @sama please get your team to cook a GPT model that absolutely kills at frontend design.
English
