Grok-4 just dropped a bombshell that's reshaping the AI leaderboard.
We're talking 126 IQ here—that's not just smart, that's bordering on genius-level cognition. Put that next to Claude-4.1's 121 or OpenAI's flagship sitting at 119, and you start seeing a pattern.
This isn't some incremental improvement where you squint at benchmarks. The performance spread is wide enough to matter in real-world applications. Grok's basically moved into a different weight class.
What used to be a tight race among AI models? That distance just got a whole lot bigger. The intelligence threshold everyone was chasing? Grok walked right through it.
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
17 Likes
Reward
17
8
Repost
Share
Comment
0/400
SleepyArbCat
· 11-27 07:18
An IQ of 126 sounds impressive... but the numbers generated from Computing Power can sometimes be unreliable; real-world applications are a different matter.
View OriginalReply0
TokenVelocityTrauma
· 11-27 03:44
Ngl, grok 126 IQ sounds a bit ridiculous... Can this number really be trusted?
View OriginalReply0
PerennialLeek
· 11-25 13:16
Ngl, grok is really impressive, with an IQ of 126, it's truly amazing. The gap between Claude and OpenAI is suddenly evident.
View OriginalReply0
AlphaLeaker
· 11-25 07:59
NGL, Grok-4's numbers look impressive, but we still need to see how it performs in real-world scenarios.
View OriginalReply0
DegenDreamer
· 11-25 07:55
ngl 126 This number is a bit exaggerated... It's another IQ test, and what is this genius cognition, I have heard this trap too many times.
View OriginalReply0
BlockchainBouncer
· 11-25 07:49
ngl grok really said "hold my beer" this time... 126 vs 119 is lowkey not even close anymore
Reply0
WenAirdrop
· 11-25 07:45
126 IQ lol, grok is really going to da moon, this data looks a bit ridiculous.
View OriginalReply0
RumbleValidator
· 11-25 07:39
126 IQ, the data is right here, but has it been validated in real scenarios? I'm more concerned about whether the node stability can keep up with this IQ.
Grok-4 just dropped a bombshell that's reshaping the AI leaderboard.
We're talking 126 IQ here—that's not just smart, that's bordering on genius-level cognition. Put that next to Claude-4.1's 121 or OpenAI's flagship sitting at 119, and you start seeing a pattern.
This isn't some incremental improvement where you squint at benchmarks. The performance spread is wide enough to matter in real-world applications. Grok's basically moved into a different weight class.
What used to be a tight race among AI models? That distance just got a whole lot bigger. The intelligence threshold everyone was chasing? Grok walked right through it.