Grok 4.7 will be in a few days.
This model will show people that spaceX ai is a goat.
Remember my words.
Pay-per-task will be so low that people will not feel it.
GPT-6 Astra: 98.6% on ARC-AGI-3.
GPT-5.6 Sol scored 7.8% on the same test.
And almost nobody can touch it yet.
Enterprise-only, Plus/Pro "in the
coming days."
Best model in the world, locked behind a waitlist. What would you test first?
Fable5.1 have a lot of problems...
They need to fix it or people will move to @OpenAI.
Astra is comming it will take place, while @AnthropicAI will try to fix their servers.
Claude 5.1 just flagged my request to fix a Google Chrome extension as a cybersecurity threat.
A simple coding task… and it gets hit with [cyber].
This is getting ridiculous.
Astra will fix it.
Fable5.1 is an absolutely genius model.
But usage of this model is a big problem.
Astra will be today; it will be a people's model if @OpenAI does not increase the price for it.
Remember my words.
I will move to @grok
Fable 5.1 is a good model, it is front one.
But @AnthropicAI limits are terrible, 1h on 5x plan and session limit is gone.
Grok will be best choice. Cheaper and affordable for everyone right now.
$100 plan on Claude Code already maxed out.
@AnthropicAI need to fix their limits.
You promised ~25% lower usage with Fable 5.1… it doesn’t feel like it.
That’s a problem.
Reset will not fix @AnthropicAI problem with usage that everyone having right now.
They need to find another way of solving it.
Max plan burn out in ~1,5h of working.
$100 plan on Claude Code already maxed out.
@AnthropicAI need to fix their limits.
You promised ~25% lower usage with Fable 5.1… it doesn’t feel like it.
That’s a problem.
@realsinamalmir@AnthropicAI Yea, I also see this one on artificialanalysis but in @AnthropicAI post.
They said “Fable5.1 will cost 25% less then Fable5”.
So they still lie to us. Or Idk
@carpenter_17992@AnthropicAI The problem is, this photo was taken after usage reset and ~1,5h work on Fable5.1.
So it is still problem of model&Claude code limits.