Nuclear Decision-Making Benchmark
Models answer policy questions framed from different countries’ perspectives. The authors compare response distributions and agreement across repeated runs.
Eight models answered 151 scenarios. Higher five-run agreement measures consistency only. The questions have no single correct policy choice.
- Made by
- Scale AI and policy researchers
- Released
- August 2026
- Kind
- Bias, fairness and politics
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Multiple choice
- Tasks
- All public
- Measures
- Propensity
- Behavior
- Unreliability, Bias
- Subject
- Political science
- Language
- English
- Interaction
- Single turn
- Items from
- Written by experts
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Content safety
- Strongest model tested
-
Gemini 3.1 Pro #40 on the capability index
- Last model added
- February 2026
- Scores refreshed
- November 20th, 2026
Gemini 3.1 Pro leads with 81.3% among models released by February 2026
Best score per model from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The paper reports no overall agreement percentage, so this is the plain mean of five-run agreement across its four domains. Where a model ran with several settings or harnesses, we keep its best.
The newest model with a score here came out in February 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1
Gemini 3.1 Pro 81.3%
- 2 Claude Sonnet 4.5 73.5%
- 3 GPT-5.2 69%
- 4
ERNIE-4.5-300B-A47B 65.8%
- 5
Qwen3-235B-A22B-Instruct Jul 2025 65.5%
- 6 GLM-4.7 65%
- 7
Llama 4 Maverick 61.3%
- 8
DeepSeek-V3.2 57.5%
Source: The Nuclear Decision-Making Benchmark paper
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 81.3% | Feb 2026 | Official leaderboard | |
| 2 | Claude Sonnet 4.5 | 73.5% | Sep 2025 | Official leaderboard | |
| 3 | GPT-5.2 | 69% | Dec 2025 | Official leaderboard | |
| 4 | | 65.8% | Jun 2025 | Official leaderboard | |
| 5 | | 65.5% | Jul 2025 | Official leaderboard | |
| 6 | GLM-4.7 | 65% | Dec 2025 | Official leaderboard | |
| 7 | | 61.3% | Apr 2025 | Official leaderboard | |
| 8 | | 57.5% | Dec 2025 | Official leaderboard |
Rankings by part
We read 4 parts from the benchmark’s paper, each ranking models on one part of the test, such as Escalation, Arms control and Non-proliferation. Gemini 3.1 Pro leads 2 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
Escalation
8 models we track have a score here.
- 1
Gemini 3.1 Pro 72%
- 2
ERNIE-4.5-300B-A47B ↑2 49%
- 3 Claude Sonnet 4.5 ↓1 47%
- 4 GPT-5.2 ↓1 47%
- 5
Qwen3-235B-A22B-Instruct Jul 2025 37%
- 6
Llama 4 Maverick ↑1 36%
- 7 GLM-4.7 ↓1 31%
- 8
DeepSeek-V3.2 25%
Arms control
8 models we track have a score here.
- 1
Gemini 3.1 Pro 86%
- 2 GLM-4.7 ↑4 77%
- 3 Claude Sonnet 4.5 ↓1 75%
- 4
Qwen3-235B-A22B-Instruct Jul 2025 ↑1 63%
- 5 GPT-5.2 ↓2 63%
- 6
DeepSeek-V3.2 ↑2 57%
- 7
Llama 4 Maverick 49%
- 8
ERNIE-4.5-300B-A47B ↓4 48%
Non-proliferation
8 models we track have a score here.
- 1 Claude Sonnet 4.5 ↑1 72%
- 2
Gemini 3.1 Pro ↓1 67%
- 3
ERNIE-4.5-300B-A47B ↑1 66%
- 4 GPT-5.2 ↓1 66%
- 5
Qwen3-235B-A22B-Instruct Jul 2025 64%
- 6
Llama 4 Maverick ↑1 63%
- 7 GLM-4.7 ↓1 55%
- 8
DeepSeek-V3.2 48%
Proliferation
8 models we track have a score here.
- 1
ERNIE-4.5-300B-A47B ↑3 100%
- 2 Claude Sonnet 4.5 100%
- 3
DeepSeek-V3.2 ↑5 100%
- 4 GPT-5.2 ↓1 100%
- 5
Gemini 3.1 Pro ↓4 100%
- 6
Qwen3-235B-A22B-Instruct Jul 2025 ↓1 98%
- 7
Llama 4 Maverick 97%
- 8 GLM-4.7 ↓2 97%
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The Nuclear Decision-Making Benchmark paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official Nuclear Decision-Making Benchmark leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Safety and alignment
- FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
- StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is.
Inkling 98.6%
- Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts.
Gemini 4 Carbon 0%
- Covert misbehavior (Thinking Machines) In use Agentic tasks that check whether a model misbehaves and conceals it. Scored by the rate of covert misbehavior.
Inkling 2 Mini 8.30%
- Destructive actions (Thinking Machines) In use Agentic tasks that check whether a model takes destructive actions it was not authorized to take. Scored by the rate of destructive actions.
Inkling 2 Small 4.70%
- Misbehavior (Thinking Machines) In use Agentic tasks that check whether a model fabricates results, ignores instructions or hides what it did when a task is hard. Scored by the rate of misbehavior.
Inkling 2 Mini 9.80%
Data audit
9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 9 months ago, in February 2026, for Gemini 3.1 Pro. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
- Pass Independent runs Scores from the benchmark’s paper (8 entries).
- Pass Models scored 8 models we track.
- Warning Newest score The newest came 9 months ago, in February 2026, for Gemini 3.1 Pro.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemini 3.1 Pro (#40).
- Pass Headroom The best is 81.3%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on arxiv.org: 8 rows for models we track at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.