Pruthvi

Made an extension to extract Instagram liked posts. It supports quick navigation & downloads the thumbnails so they are always present.

Just 6 month ago, OpenAI had lost it's top position Users were switching to Gemini Opus 4.5 was launched, miles better than any OAI model Now, Google is doing it's best to catch-up Anthropic is busy doing its own bullshit OpenAI is at top again. Xai is nowhere to be seen.

@vansh22b What's the cost breakdown? what did they (hackers I presume) use mostly? I remember @DhravyaShah got some 40k $ something API cost from X after price changes, he just sold that biz.

It's def. not easy to be me.

I supported ChatGPT and OpenAI while the competitors were firing shots from every direction. I always backed you to be the best. I was one of the earliest investors, at least in spirit. I even bought the Go plan. Through outages, rate limits, and every unexpected response, I stayed loyal and believed the future belonged to you. I never used Cursor. I never used Claude. I stayed faithful to ChatGPT and Codex. While others wandered off to try every new AI that appeared on the internet, I stood my ground. I defended OpenAI in debates nobody asked me to join. My loyalty was unquestionable and my commitment was absolute. But this is where I draw the line. A Messi advertisement was apparently the final test and I have failed it. After everything we have been through, this is a betrayal I cannot overlook. Effective immediately, I am officially Team Anthropic. I hereby renounce my allegiance and declare a new era. CLAUDE AI IS THE GREATEST THERE WAS THE GREATEST THERE IS THE GREATEST THERE EVER WILL BE.

@notjazii They are for enterprises mostly

Is Claude Fable 5 actually too expensive? ran a small test to see how bad it is > bought Hermes Plus and got $22 worth of credits > Fable burned $20 in 30 minutes with just 2 prompts now this might not feel the same for claude plan users because plan limits work differently so use it while you can till June 22 after that, pay-per-use is gonna get way more expensive would you still use Fable daily at pay-per-use pricing?

@matthewmillerai let's be honest nobody used Opus 4.1 it was not that big of a leap, but fable 5 is.

Claude Fable 5 is 2x the cold of Claude Opus 4.8. People are complaining about it, but don't forget Claude Opus 4.1.

@xikhar @LexnLin I ain't believing this until someone replicates it

@LexnLin Try replicating this x.com/i/status/20644…

Finally home from school. Will share prompts and continue fabblemax :) If you got interesting prompts to test, reply below with them!

@ChiragAgg5k I believe there's a Github feature exactly for this

someone keeps raising random prs in appwrite repo playing with agents is fine and all but can you please do it in your own repos?

@arnesh_11 It's cause they don't want you to do cybersecurity stuff with it. It is found to be too good at that, and poses huge risk of finding vulnerabilities in a lot of widely used softwares

All I did was ask it to audit my codebase man what 😭

@sankitdev @capcutapp opencut.app

someone was building opensource of @capcutapp. is this done???

@theo Can we get Quipslop live for sometime again The new Fable 5, GPT 5.5, Kimi 2.6, Minimax M3, Muse Spark, Deepseek v4 Pro will definitely show some leaderboard changes

@fedethecreator1 @nicdunz this one most likely discord.com/invite/Mk298GP…

@nicdunz what that discord brother

claude fable 5 just decimated the competition

Did not expect codex to throw its system prompt this easily

@ayushunleashed @abrilzucchi I thought you had a job :?

@abrilzucchi 0/4 🙌

no person on this app has all 4: - a full time job - social life - mentally stable - a serious relationship

@mememandir Always make sure to see the pricing on checkout page

i had to sign a contract urgently so proceeded to edit with this website- it was the first search result & it was “free”. the contract was so fucking long that when i reached the end, it was already too late to cancel. it was a 50₹ charge. i ended up paying with a card dont think they had upi and guess what? these fuckers created a mandate with my card to charge the same amount every month

@Jeremy_ @LexnLin 50% up from last year :)

@pruthvithinks @LexnLin Oh my bad a Gemini distillation. Apple is still beyond cooked.

Apple Intelligence

@Jeremy_ @LexnLin They bought a model from google. They OWN it.

@LexnLin Gemini wrapper?

@krishdotdev replace NVIDIA with xAI

It's not FAANG anymore. It's MANGO.

@SinghDevHub @adxtyahq @NamanMarkh11301 If you are programming stick to the US models

@adxtyahq @NamanMarkh11301 Is it at par with the claude subsidized 100$ plan ?

after using nvidia nemotron 550B for several hours: we need fewer benchmark screenshots and more people actually using the model before declaring it the future

@adxtyahq @NamanMarkh11301 because it thinks for unreasonably long

@NamanMarkh11301 I have been using it internally, its cheap and really good but takes a lot of time (the only issue i feel)