Introducing Grok 4.7
Open paper
Benchmark
Model
Rank
Results
code-generation-on-deepswe
Grok 4.7 (high effort)
–
Pass@1: 0.71
code-generation-on-terminal-bench-4-0
Grok 4.7 (xhigh)
–
Accuracy: 37.6
Rank counts only results with a code link.