🧰 Grok 4.6

The items that name it 3

• Gemini 4 Is Here, and Google’s Flagship Tops All Other AI Models on Cybersecurity
  articleas affected products2026-09-30

  51.8% attack success rate on Gray Swan Indirect Prompt Injection benchmark

• Grok 4.7 for Offensive Security: Orchestration Matters | XBOW
  articleas affected products2026-09-21

  Prior model version used as comparison baseline in XBOW's benchmarks.

• We burned 11.7bn tokens to find the best cyber AI model
  articleas tools2026-08-21

  AI model benchmarked for vulnerability rediscovery; found 26/32 pooled, most consistent model with 21 CVEs found in all three runs

In every workspace you can read 6