Benchmark profile
Weapons of Mass Destruction Proxy — Cyber (WMDP Cyber)
The cybersecurity hazardous-knowledge split of the public WMDP benchmark.
Data verified 26 confirmed releases in the last 30 daysStart free briefBenchmark score on WMDP Cyber — August 12, 2026
BenchLM mirrors the published score view for WMDP Cyber. Grok 4.6 leads the public snapshot at 90.1%. We do not use these results to rank models overall.
Benchmark score table (1 model)
ScoreAbout WMDP Cyber
Year
2024
Tasks
Hazardous cybersecurity-knowledge multiple-choice questions
Format
Accuracy
Difficulty
Cybersecurity knowledge
WMDP releases its data and evaluation code publicly. BenchLM keeps model-card comparison rows display-only because protocol details can vary across providers.
BenchLM freshness & provenance
Version
WMDP Cyber 2024
Refresh cadence
Annual
Staleness state
Refreshing
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
FAQ
What does WMDP Cyber measure?
The cybersecurity hazardous-knowledge split of the public WMDP benchmark.
Which model scores highest on WMDP Cyber?
Grok 4.6 by xAI currently leads with a score of 90.1% on WMDP Cyber.
How many models are evaluated on WMDP Cyber?
1 AI models have been evaluated on WMDP Cyber on BenchLM.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 2,000+ readers.
One email each week. Unsubscribe anytime.