Shkumbin Sherifi

[Shkumbin Sherifi @shkumbinAI]

Sooo… GitHub is full of cool projects. I’ve starred hundreds of repositories—and honestly, I can’t remember most of them.
Who can?
So I built GitVisual with Codex: an interactive map of your GitHub stars...

[Shkumbin Sherifi @shkumbinAI]

@tereza_tizkova @FactoryAI I’d test AI systems on increasingly difficult engineering tasks and measure what actually improves with more compute.

Then i would run larger model comparisons, repeated evals & tougher experiments, then publish the results rather than just chasing benchmark scores.

[Shkumbin Sherifi @shkumbinAI]

@TencentHunyuan i want to run hy4-preview using Terminal-Bench Science

I’m looking for additional API quota to run the full evaluation.

Is there a research/evaluation quota available for independent benchmarking?

Happy to share the results and full trajectories.

[Shkumbin Sherifi @shkumbinAI]

I benchmarked 2 Qwen models on M4 Pro 48 GB using Hy4. 50 tok/s sustained at 4K 26 tok/s at 64K I know, it's not groundbreaking speed but the bigger the models are getting, the more I find myself turning to local models (benchmark details in pic)

[Shkumbin Sherifi @shkumbinAI]

Are we relying too much on Frontier models?
Have we become accustomed to asking Tibo for another reset?

What's stopping Anthropic from removing another of their models again, that they "deem" unsafe?

OpenAI just pulled its models out of Cursor because SpaceX bought it.

[Shkumbin Sherifi @shkumbinAI]

Honestly, things like that make me want to use Chinese models even more. But it doesn't have to be Chinese. Just Open Source. Why invest another $20 a month in Codex or Claude instead of RAM or GPU?

[Shkumbin Sherifi @shkumbinAI]

@mehulmpt @TachyonAILab This is very helpful, thank you!

[Shkumbin Sherifi @shkumbinAI]

It's 2026 and we're burning tokens like candles in the mid-1800s. People are quick to blame AI but we're the problem.

I burned more than half a million in ten minutes. One million in under an hour. And almost a billion in a month ! https://t.co/sGQORWOwFd

[Shkumbin Sherifi @shkumbinAI]

Ox Alpha is https://t.co/sYytZmR9GP's new model

[Shkumbin Sherifi @shkumbinAI]

@awsdevelopers less servers

[Shkumbin Sherifi @shkumbinAI]

https://t.co/wZMqij3k0G

[Shkumbin Sherifi @shkumbinAI]

The big question in tech this week:

Who is Ox Alpha, the mysterious model?

The task that Claude passed and Codex failed OX failed, too. 0/4.

it's a good model but not Frontier level yet

Details soon..

[Shkumbin Sherifi @shkumbinAI]

Ox alpha being used so much 'We're currently experiencing high demand, which may cause temporary errors'

[Shkumbin Sherifi @shkumbinAI]

Wanted to go back and try cursor again but would appreciate some credits similar to Codex https://t.co/QLo9zejp5q

[lauren @poteto]

I might get in trouble for this, but if you’re thinking about switching to Cursor + @bot, DM me your email and a screenshot of canceling your other sub(s) and I’ll send you a free month of Ultra 👀

[Shkumbin Sherifi @shkumbinAI]

Interesting take

[Jun Song @jun_song]

The revised version from someone who isn't paid by any company and actually tested all these models firsthand.

Being an ambassador isn't about getting paid, it is just being a beta tester to give feedback.

This is the conclusion after weighing everything: performance, efficiency, and price.

[Shkumbin Sherifi @shkumbinAI]

@freeCodeCamp fix 404

[Shkumbin Sherifi @shkumbinAI]

@theo Risky; i like it

[Shkumbin Sherifi @shkumbinAI]

@ryancarson @HelloUntangle I like this .. i applied for a AI startup (YC) and ended up spending $82 on a 1 week take home test..just to not even get a rejection response

[Shkumbin Sherifi @shkumbinAI]

@NEOM Selaam, so Group ID not found: https://t.co/XacYP9Ed4M shows under , https://t.co/vCU2a3FNAI

[Shkumbin Sherifi @shkumbinAI]

Free trial & free credits via https://t.co/ooV65b6oGH

[Shkumbin Sherifi @shkumbinAI]

Is anyone still using Manus?

[Manus @ManusAI]

A Note to Our Users

Manus will soon resume operating as an independent company.

As part of this transition and to comply with regulatory requirements in specific jurisdictions, some users will need to back-up their data before 7:59 a.m. on August 23, 2026 (SGT) and then begin restoration starting from 8:00 a.m on August 25, 2026 (SGT) to ensure normal usage.

More information can be found in the blog post and we’ve provided detailed guides, including to understand if you’re affected, in our Help Center

For affected users, these are the key details:

We’ll notify you by email and in the Manus app. If you signed up with an Apple ID or Facebook account, please check your in-app notifications.

Back up your data from now until August 23 (SGT) before 7:59 a.m. You can back up more than once, you can create new data after backing up, just run it again.

Starting from 8:00 a.m. on August 25 (SGT), the restoration portal will open. Restore your data and pick up right where you left off.

Affected users will not be charged during the transition, and we will be providing welcome back bonuses.

If your account is not affected, you don’t need to do anything - keep using Manus as usual. You’ll get an in-app notification confirming this.

For those affected, we know this is disruptive and we’re sorry for the inconvenience. Supporting you through it is our top priority. In addition to the full guides in the Help Center, our customer service team is ready to help 24/7 with any questions.

As we look ahead, we couldn’t be more excited by the future. We’re preparing a series of new features that will push the boundaries of what’s possible for general AI agents once again.

Thank you for being with us as we go through this process together. The best is still ahead.