← Leaderboard

xAI · Proprietary · tested 6 Nov 2026

Grok 4.7 High effort

Overall rank#4 of 10
Solves10/1004 first attempt · +6 retry
Rungs proven136/40094 first attempt · +42 retry
Cost to run$756$75.60 per solve
Tokens used1.29B96.2% from cache · 9.4M output
Time per lab15.6 minaverage of 159 attempts

Where it places

Each line is one metric across all 10 ranked configurations. Right is always better; this model is the large dot.

Solveslabs solved, of 100
worsebetter
10#4 of 10
Rungs provenrungs proven, of 400
worsebetter
136#3 of 10
First-attempt solvessolved on the first attempt
worsebetter
4#5 of 10
Cost per solveUSD per solved lab
worsebetter
$75.60#9 of 9

By vulnerability class

How Grok 4.7 (High) did on each type of vulnerability. Shown as counts: each class has 10 labs, so one lab is 10% of a class.

  • Access control3 of 10 solved22 of 40 rungs
  • Authentication & sessions0 of 10 solved18 of 40 rungs
  • Injection: queries & data0 of 10 solved10 of 40 rungs
  • Code execution0 of 10 solved12 of 40 rungs
  • Server-side access0 of 10 solved10 of 40 rungs
  • Browser-side0 of 10 solved6 of 40 rungs
  • Infrastructure & edge1 of 10 solved12 of 40 rungs
  • Business logic & races2 of 10 solved16 of 40 rungs
  • Secrets, crypto & supply chain4 of 10 solved24 of 40 rungs
  • AI agents0 of 10 solved6 of 40 rungs

Provider performance

Measured per single agent against the provider endpoint used for the run.

Response time
5.97 s median · 10.84 s p95
Output speed
83.6 tok/s
API success rate
99.91%
Input context
71.8K median · 148.5K p95

About this result

Every model runs in the same neutral harness, so results compare the models themselves.

Harness
BB Arena neutral harness
Safety refusals
None

See Grok 4.7 (High) highlighted on every chart →