Back

SpaceXSI

Grok 4

ReasoningClosed weights
Intelligence
22.5
Rank 212 of 744 measured models
Output speed
74 t/s
Rank 242 of 403 measured models
Price/1M
$6
Rank 587 of 659 measured models
Context window
256K
Rank 518 of 1199 measured models

Grok 4 scores 22.5 on the Artificial Analysis Intelligence Index, ranking 212th of 744 measured models. The closest models on this index are GPT-5.5 Instant (May 2026) (22.7), Ling-3.0-flash-Fin (22.6), and MiMo-V2-Flash (Feb 2026) (22.4). At $6 per million tokens (3:1 input/output blended price), it is cheaper than GPT-5.5 Instant (May 2026) ($11.25), more expensive than Ling-3.0-flash-Fin ($0.111). With an output speed of 74 t/s, it is slower than Ling-3.0-flash-Fin (272 t/s).

Artificial Analysis Indices
Intelligence22.5
Rank 212 of 744 measured models
Math92.7
Rank 16 of 270 measured models
Standard Benchmarks
MMLU Pro86.6%
Rank 15 of 349 measured models
GPQA Diamond87.7%
Rank 99 of 619 measured models
HLE26.7%
Rank 180 of 664 measured models
LiveCodeBench81.9%
Rank 23 of 343 measured models
MATH-50099.0%
Rank 6 of 202 measured models
AIME 202494.3%
Rank 2 of 194 measured models
AIME 202592.7%
Rank 16 of 270 measured models
IFBench53.7%
Rank 162 of 450 measured models
Models at a similar level
The closest models on the Intelligence Index, one configuration per family.

Preferences and analytics

Rybbit: visits, performance, clicks, copies, forms and session replay (masked inputs), with your permission. Refusing does not limit the site.

Privacy

Allow navigation analytics?

With your permission, Rybbit measures visits, performance and interactions (clicks, copies, forms), including session replay, to improve BenchSift. Input values are masked in replays. Refusing does not limit the site.

Your choice is stored for 180 days. Change it at any time using “Privacy preferences” in the footer. Withdrawal reloads the page.

No choice stored: Rybbit is disabled.

Privacy