Aflan

Al News and Benchmarks

[Aflan @Aflanai]

@Mr_nexp Can't wait to test it

[Aflan @Aflanai]

OpenAI is starting to lose the momentum it once had.

Almost two months without a new Astra model, ChatGPT and Work usage limits are being merged, Dots is facing backlash, and app bugs keep piling up.

Even the usual excitement around upcoming model releases seems unusually quiet.

OpenAI used to set the pace for everyone else. Now Google and Anthropic are stealing the spotlight.

[Aflan @Aflanai]

I'm not waiting for anything else this week except GPT-6.2 Astra.

OpenAI needs to release it, challenge Fable 5.2, and push Anthropic to bring Fable 5.5 forward. That's how we get faster progress from each side.

Fable 5.2 and Qwen 4 have already dropped this week. Meanwhile OpenAI is still holding back.

Almost two months and half without a new frontier release from Openai. What are they waiting for?

[Aflan @Aflanai]

Looks like Gemini 4 Hydrogen might be next.

Google has been releasing models at an insane pace, and they still have more lined up.

The Gemini team is showing no signs of slowing down. 🔥

[Aflan @Aflanai]

🚨 Thursday could finally be GPT-6.2 day. Start saving those banked resets because the big ship might be arriving!

What an insane week for AI: a new Gemini today, Fable 5.2 yesterday, GPT-6.2 Astra potentially tomorrow, and Qwen 4 in the mix too.

[Aflan @Aflanai]

@Mr_nexp Release the f the model

[Aflan @Aflanai]

🚨 GPT-6.2 Astra might be much closer than we think.

GPT-6.2 is now appearing in OpenAI's official documentation, with the phrase “On models 6.2 and later.”

this could point to an entire 6.2 lineup: Astra, Sol, Luna and Terra.

[Aflan @Aflanai]

@OpenAI Anything but a new model eh?

[Aflan @Aflanai]

OpenAI, JUST DROP GPT-6.2 ASTRA ALREADY 😭LOCK IN. 💀

[OpenAI @OpenAI]

We're turning our lessons from defending OpenAI into practical tools for our customers. On Thursday, join us at Cyber at the Frontier, a live keynote, for our latest cyber announcements.

Working with our frontier models, we've identified and resolved important vulnerabilities in our own code and infrastructure. That experience is informing the tools and playbooks we're building to help other security teams strengthen their defenses.

@gdb shares what we've learned and why security teams should keep putting the latest models to work as capabilities improve.

Tune in on November 12 at 10 a.m. PT: https://t.co/jYly2SEDM0

[Aflan @Aflanai]

🚨 OpenAI has a massive opportunity to make a comeback.

Opus 5.5 has been losing momentum, while Fable 5.2 burns tokens like crazy

If OpenAI delivers a faster, more efficient model with the next-generation Sol, they could turn the tables.

The timing couldn't be better.

[Aflan @Aflanai]

Fable 5.2 might have just triggered a chain reaction across the entire AI industry.

OpenAI could forced with GPT-6.2 Astra or GPT-6 Bel.

Google may forced Gemini 4 Krypton.

SpaceX could forced with Grok 4.8.

Meanwhile, Anthropic might already be preparing Opus 5.6 and Fable 5.5 as its next moves.

Every lab has an incentive to respond, and we could be heading into one of the biggest model release waves yet.

The next few weeks are going to be INSANE.

[Aflan @Aflanai]

OpenAI seriously needs to wake up and ship GPT-6.2 Astra and Sol.

Anthropic and Google are pulling ahead, and OpenAI is starting to look like the third-place lab. That's not where they should be.

Enough delays. It's time to show what they've been working on.

just release something already. 😭

[Aflan @Aflanai]

🚨 OpenAI's next releases maybe getting very close.

GPT-6.2 Astra appears to have cleared its testing phase after the earlier issues, while GPT-6.2 Sol is now in the final stages of testing.

Both models could launch soon.

But there's more: another major model is already being trained, potentially targeting a late December or early January release.

OpenAI's roadmap is looking pretty packed. Hopefully, nothing gets delayed again.

[Aflan @Aflanai]

@Mr_nexp Google is officially back

[Aflan @Aflanai]

🚨 GPT-6.2 Sol might be one of the biggest upgrades Plus users have received.

The difference is noticeable: stronger performance without burning through usage nearly as quickly.

OpenAI is finally making the $20 Plus plan feel much more worthwhile.

[Aflan @Aflanai]

If OpenAI introduces a $50 Pro plan, I hope they also double the usage limits for Plus.

Millions of users rely on the $20 subscription because they can't afford to pay more. They deserve better limits and new features too.

Don't leave Plus users behind. 👀

[Aflan @Aflanai]

The $50 ChatGPT Pro plan could be a game changer for Plus users.

5× higher usage limits, no restrictive 5-hour window, access to top-tier models and Dots.

Not everyone can afford $100/month for Pro, but $20 Plus can feel limiting if you use ChatGPT heavily.

A $50 option would finally give users a proper middle ground between the two plans.

OpenAI, make it happen. 👀

[Aflan @Aflanai]

The $50 ChatGPT Pro plan could be a game changer for Plus users.

5× higher usage limits, no restrictive 5-hour window, access to top-tier models and Dots.

Not everyone can afford $100/month for Pro, but $20 Plus can feel limiting if you use ChatGPT heavily.

A $50 option would finally give users a proper middle ground between the two plans.

OpenAI, make it happen. 👀

[Tibo @thsottiaux]

Would a Pro 50 make sense? Wdyt?

[Aflan @Aflanai]

@Mr_nexp Performance drops on current tools are not acceptable

[Aflan @Aflanai]

I've noticed GPT-6.1 Terra becoming noticeably slower, while GPT-6 Astra's overall output quality seems to have taken a hit as well.

OpenAI shouldn't have to sacrifice the performance of its existing models just to make room for new releases.

Users are paying for these models right now, not just waiting for whatever comes next.

If this is a trade-off happening behind the scenes, it's genuinely disappointing. New models should improve the experience, not make the current lineup worse. 😕

[Aflan @Aflanai]

🚨 There may be a reason OpenAI pulled the brakes on GPT-6.2 Astra and Sol.

I'm hearing that both models encountered some alignment and safety problems during final testing, potentially pushing the launch back by a few days.

But the early access feedback sounds very promising:

→ Performance is impressing testers.
→ Speed is getting positive reactions.
→ Pricing is looking competitive.

If these reports are accurate, the models themselves seem to be in good shape despite the last-minute setbacks.

Take the extra time, OpenAI. Just make sure Astra and Sol live up to the hype.