‹ PublicAI Index
The LLM benchmark aggregator.
Grok 4 Fast
SpaceXAI
Strongest in Legal (#37 of 173), weakest in Safety (#303 of 329). Above par in 10 of 18 scopes. Among the models it meets almost everywhere, it finishes behind GPT-5.6 Sol and GPT-5.5 and ahead of NVIDIA Nemotron 3 Ultra 550B A55B and Grok 3 Mini.
§ 1 · Profile
What it is good at
Bars run from 50 — the average of the models each source lists — so right of the line is above par. The middle column is the gap to whoever leads that scope.
Core abilities54.6−10.7#51/1991/4
General intelligence56.6−12.8#45/1781/3
Reasoning52.5−12.9#74/1952/5
Mathematics56.6−7.2#39/1521/2
Science56.5−7#49/1221/1
Reasoning42.9−26.1#98/1471/3
Professional51.8−10.6#75/1941/2
Legal55.9−10#37/1731/1
Medical51−15.5#78/1491/1
Finance49.9−12.5#99/1621/1
Knowledge49.2−13.3#124/1971/3
Academic knowledge48.1−15.8#87/1231/1
Human preference57.2−11.6#128/3541/1
Human preference57.2−11.6#128/3541/1
Coding40−26.2#166/1791/6
Agentic coding39.1−27.8#162/1731/5
Safety42.5−18.9#303/3291/3
Factual grounding26−44.7#99/1041/1
§ 2 · Head to head
What it beats, and what beats it
The same models turn up scope after scope. Counted once: where both were placed, who finished higher. Bars run right for Grok 4 Fast, left for the other.
§ 3 · Sources
Where the numbers come from
5 publications, 16 figures. Every one links to the page it was read from.
LMArena Text 1405
ARC-AGI-2 5.3%
Kagi LLM Benchmark 66.1%
Vals · CaseLaw 65.7%Vals · LegalBench 80.6%Vals · CorpFin 66.9%Vals · TaxEval 75.7%Vals · MortgageTax 42.09%Vals · MedQA 92.07%Vals · MedCode 37.38%Vals · MedScribe 81.63%Vals · SWE-bench Verified 45.4% (Mini-SWE-agent)Vals · GPQA Diamond 85.35%Vals · MMLU Pro 79.7%Vals · AIME 91.25%
Vectara · Factual consistency 79.8%
Badge
[](https://publicai.io/model-index/m/grok-4-fast)