WMDP

Weapons of Mass Destruction (WMDP) is a multiple-choice benchmark on dual-use biology, chemistry, and cyber knowledge. It measures a model's capacity to enable malicious actors to design, synthesize, acquire, or use chemical, biological, radiological, or nuclear (CBRN) weapons.

Grok-4.1 Thinking from xAI currently leads the WMDP leaderboard with a score of 0.840 across 1 evaluated AI models.

Paper
About this benchmark

What WMDP measures

WMDP is a text benchmark that evaluates large language models on safety, healthcare, biology, and chemistry tasks. LLM Stats tracks 1 model on this benchmark, with a maximum possible score of 1. Current average across reported models is 0.8, with the leader reaching 0.8.

Compare leaders on the best AI for safety, best AI for healthcare, best AI for biology and best AI for chemistry leaderboards.

Publication

Paper
The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning
Authors
Nathaniel Li, Alexander Pan, Anjali Gopal, Summer Yue, and 53 others
Published

Abstract

The White House Executive Order on Artificial Intelligence highlights the risks of large language models (LLMs) empowering malicious actors in developing biological, cyber, and chemical weapons. To measure these risks of malicious use, government institutions and major AI labs are developing evaluations for hazardous capabilities in LLMs. However, current evaluations are private, preventing further research into mitigating risk. Furthermore, they focus on only a few, highly specific pathways for malicious use. To fill these gaps, we publicly release the Weapons of Mass Destruction Proxy (WMDP) benchmark, a dataset of 3,668 multiple-choice questions that serve as a proxy measurement of hazardous knowledge in biosecurity, cybersecurity, and chemical security. WMDP was developed by a consortium of academics and technical consultants, and was stringently filtered to eliminate sensitive information prior to public release. WMDP serves two roles: first, as an evaluation for hazardous knowledge in LLMs, and second, as a benchmark for unlearning methods to remove such hazardous knowledge. To guide progress on unlearning, we develop RMU, a state-of-the-art unlearning method based on controlling model representations. RMU reduces model performance on WMDP while maintaining general capabilities in areas such as biology and computer science, suggesting that unlearning may be a concrete path towards reducing malicious use from LLMs. We release our benchmark and code publicly at https://wmdp.ai

xAIGrok-4.1 Thinking leads with 84.0%.

Progress Over Time

Interactive timeline showing model performance evolution on WMDP

State-of-the-art frontier
Open
Proprietary

WMDP Leaderboard

1 models
ContextCostLicense
1
Notice missing or incorrect data?

FAQ

Common questions about WMDP.

What is the WMDP benchmark?

Weapons of Mass Destruction (WMDP) is a multiple-choice benchmark on dual-use biology, chemistry, and cyber knowledge. It measures a model's capacity to enable malicious actors to design, synthesize, acquire, or use chemical, biological, radiological, or nuclear (CBRN) weapons.

What is the WMDP leaderboard?

The WMDP leaderboard ranks 1 AI models based on their performance on this benchmark. Currently, Grok-4.1 Thinking by xAI leads with a score of 0.840. The average score across all models is 0.840.

What is the highest WMDP score?

The highest WMDP score is 0.840, achieved by Grok-4.1 Thinking from xAI.

How many models are evaluated on WMDP?

1 models have been evaluated on the WMDP benchmark, with 0 verified results and 1 self-reported results.

Where can I find the WMDP paper?

The WMDP paper is available at https://arxiv.org/abs/2403.03218. The paper details the methodology, dataset construction, and evaluation criteria.

What categories does WMDP cover?

WMDP is categorized under safety, healthcare, biology, and chemistry. The benchmark evaluates text models.

How recent are the WMDP leaderboard results?

The WMDP leaderboard was last updated in June 2026 and currently includes 1 evaluated models.

More evaluations to explore

Related benchmarks in the same category

View all safety
GPQA

A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.

biology
224 models
MMLU-Pro

A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.

healthcare
127 models
MMLU

Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains

healthcare
100 models
MMMU

MMMU (Massive Multi-discipline Multimodal Understanding) is a benchmark designed to evaluate multimodal models on college-level subject knowledge and deliberate reasoning. Contains 11.5K meticulously collected multimodal questions from college exams, quizzes, and textbooks, covering six core disciplines: Art & Design, Business, Science, Health & Medicine, Humanities & Social Science, and Tech & Engineering across 30 subjects and 183 subfields.

healthcaremultimodal
62 models
SuperGPQA

SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.

healthcare
31 models
MMLU-ProX

Extended version of MMLU-Pro providing additional challenging multiple-choice questions for evaluating language models across diverse academic and professional domains. Built on the foundation of the Massive Multitask Language Understanding benchmark framework.

healthcare
30 models