Skip to content

xAI

Grok 4.7

xAI's latest model. Grok 5 has not shipped: xAI's release notes end at 4.7.

Announced
Access
Public

Benchmark scores

TestScoreBest on the trackerHow it was measured
Epoch Capabilities Indexmany tests combined into one score156 pointsGPT-6.2 Astra 174 pointsIndependentlisted as "Grok 4.7", CI 154.15 to 158.13Run 9 Nov 2026
AA Intelligence Index v5about ten tests in one score35.5 pointsClaude Fable 5.2 64.6 pointsIndependentv5.0.1; listed as Grok 4.7 (Xhigh)Run 10 Nov 2026
FrontierSWE V2very long engineering jobs30%Claude Fable 5.2 70%Independentproximus harness, xhigh effort, mean@5 over 34 tasks, 20-hour budget (best@5 41.1)
Terminal-Bench 4.1long command-line tasks39%Claude Opus 5.5 66%IndependentGrok Build harness, xhigh effort, 330 trials, ±3.59 (95% CI)
APEX-Agentsprofessional office tasks55%Gemini 4 Carbon 84%Independentpass@1, unknown effort; listed as "Grok 4.7"
ARC-AGI-3puzzle games, standard harnessTypical first-time human player: 100%1.8%Claude Fable 5.2 81%IndependentARC Prize config "Grok 4.7 (XHigh)" (xai-grok-4-7-xhigh), standard harness
FrontierMath T1–3hard maths problems53%Claude Fable 5.2 95%Independentxhigh reasoning effort (grok-4.7_xhigh)Run 22 Sept 2026
GPQA DiamondPhD-level science questionsPhD experts: 70%Saturated93%Gemini 4 Argon 96%Independentxhigh reasoning effort (grok-4.7_xhigh)Run 22 Sept 2026
FrontierMath T4research-level maths problemsSaturated17%GPT-6.1 Sol 100%Independentxhigh reasoning effort (grok-4.7_xhigh)Run 22 Sept 2026
ARC-AGI-2visual puzzles, model aloneAverage person: 60%Saturated61%GPT-6 Astra 95%IndependentARC Prize config "Grok 4.7 (XHigh)" (xai-grok-4-7-xhigh)
OTIS Mock AIMEcompetition mathsSaturated98%GPT-5.5 100%Independentxhigh reasoning effort (grok-4.7_xhigh)Run 22 Sept 2026

Compare every model on the same tests →

Sources