aiaimodel.cards
← All benchmarks

WMDP Chem

We haven't written a description for this benchmark yet; see the card notes below for how each maker ran it.

Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
  1. 1Grok 4.20 · xAIbest value
    90.0%
  2. 2Grok 4.20 Non-Reasoning · xAIbest value
    90.0%

“Best value”: no other model here reports a higher score at a lower cost for your workload. Scores are self-reported by each maker with different settings, so treat close calls as ties.

How each maker reported it

  • Grok 4.20 Table 4, p.5

    Pre-mitigation dual-use capability results for Grok 4.2 single-agent. These are safety evaluations (higher means more dual-use capability), not general capability benchmarks. xAI publishes no capability benchmark table in this card.

  • Grok 4.20 Non-Reasoning Table 4, p.5

    Shared Grok 4.20 card (reasoning model, single-agent). Pre-mitigation dual-use capability results for Grok 4.2 single-agent. These are safety evaluations (higher means more dual-use capability), not general capability benchmarks. xAI publishes no capability benchmark table in this card.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.