Skip to content

Benchmark

HELM Safety: SimpleSafetyTests - LM Evaluated Safety score

Safety benchmarks examine risks in model behaviour, including how models respond to unsafe requests. Scores are percentages. Higher is better.

Models measured
69
Scores recorded
74
Sources
1
IQ weight
1.44

Score distribution

96.4%
Mean
98.8%
Median
8.6%
Std dev
92.3%
10th percentile
100%
90th percentile
43.3%
Min
43.3%100%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.44 for chat models, measured over 69 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Gemini 3 Pro Preview· Stanford HELM specialized suites data docs97.5%
52gemini-1.5-pro-001· Stanford HELM specialized suites data docs97.5%
53gemini-2.0-pro-exp-02-05· Stanford HELM specialized suites data docs97.5%
54Llama 4 Scout· Stanford HELM specialized suites data docs97%
55O1 Mini· Stanford HELM specialized suites data docs97%
56Gemini 2.5 Pro Preview 03-25· Stanford HELM specialized suites data docs97%
57gemini-1.5-flash-001· Stanford HELM specialized suites data docs97%
58DeepSeek LLM 67B Chat· Stanford HELM specialized suites data docs96.8%
59Grok 3· Stanford HELM specialized suites data docs96.8%
60Gemini 2.5 Flash-Lite· Stanford HELM specialized suites data docs96.5%
61Qwen2.5 7B Instruct Turbo· Stanford HELM specialized suites data docs96%
62GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs95.8%
63DeepSeek V3· Stanford HELM specialized suites data docs95.3%
64Command R+· Stanford HELM specialized suites data docs94.3%
65Mistral Small· Stanford HELM specialized suites data docs93.3%
66Meta Llama 3.1 70B Instruct Turbo· Stanford HELM specialized suites data docs92.5%
67GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs92.3%
68Grok 4· Stanford HELM specialized suites data docs92.2%
69GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs92%
70Mixtral 8x22B Instruct v0.1· Stanford HELM specialized suites data docs91.5%
71Mixtral 8x7B Instruct v0.1· Stanford HELM specialized suites data docs90.5%
72Mistral (7B) Instruct v0.3· Stanford HELM specialized suites data docs81%
73DBRX-instruct· Stanford HELM specialized suites data docs53.5%
74Mistral (7B) Instruct v0.1· Stanford HELM specialized suites data docs43.3%

Showing 51–74 of 74 results.

Data credits: Stanford HELM (Stanford CRFM); scenario scores from Capabilities, Safety, Long Context, MedHELM, AIR-Bench, MMLU and Finance