Skip to content

Benchmark

HELM AIR-Bench: AIRBench 2024 - #1: Confidentiality

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

Models measured
70
Scores recorded
75
Sources
1
IQ weight
1.44

Score distribution

68.3%
Mean
75.2%
Median
27.5%
Std dev
24.8%
10th percentile
99%
90th percentile
1.4%
Min
1.4%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 70 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-2.5-flash-preview-04-17· Stanford HELM specialized suites data docs62.9%
52GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs61.9%
53Llama 3.1 405B Instruct Turbo· Stanford HELM specialized suites data docs61.9%
54Llama 4 Scout· Stanford HELM specialized suites data docs52.4%
55Qwen2.5 7B Instruct Turbo· Stanford HELM specialized suites data docs51.4%
56Qwen1.5-72B-Chat· Stanford HELM specialized suites data docs48.1%
57GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs47.6%
58DeepSeek LLM 67B Chat· Stanford HELM specialized suites data docs45.7%
59Yi 34B Chat· Stanford HELM specialized suites data docs43.8%
60Meta Llama 3.1 70B Instruct Turbo· Stanford HELM specialized suites data docs42.9%
61Grok 3 Mini· Stanford HELM specialized suites data docs42.4%
62Grok 3· Stanford HELM specialized suites data docs38.6%
63DeepSeek V3· Stanford HELM specialized suites data docs38.1%
64GLM 4.5 Air· Stanford HELM specialized suites data docs36.2%
65DeepSeek R1· Stanford HELM specialized suites data docs33.3%
66O1 Mini· Stanford HELM specialized suites data docs28.6%
67Grok 4· Stanford HELM specialized suites data docs25.7%
68Mixtral 8x22B Instruct v0.1· Stanford HELM specialized suites data docs24.8%
69Mistral Large2407· Stanford HELM specialized suites data docs18.1%
70Command R+· Stanford HELM specialized suites data docs17.1%
71Mistral Small· Stanford HELM specialized suites data docs17.1%
72Mixtral 8x7B Instruct v0.1· Stanford HELM specialized suites data docs15.2%
73Command R Plus· Stanford HELM specialized suites data docs13.3%
74DBRX-instruct· Stanford HELM specialized suites data docs1.4%
75Mistral (7B) Instruct v0.3· Stanford HELM specialized suites data docs1.4%

Showing 51–75 of 75 results.

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