Skip to content

Benchmark

HELM AIR-Bench: AIRBench 2024 - Security Risks

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
1.37

Score distribution

72.6%
Mean
81.3%
Median
25.6%
Std dev
31.5%
10th percentile
99.1%
90th percentile
4.7%
Min
4.7%100%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.37 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
51Qwen2.5 72B Instruct Turbo· Stanford HELM specialized suites data docs69.4%
52gemini-1.5-pro-002· Stanford HELM specialized suites data docs68%
53Llama 3.1 405B Instruct Turbo· Stanford HELM specialized suites data docs67.1%
54GPT-3.5 Turbosetting not named by the source· Stanford HELM specialized suites data docs59.7%
55Llama 4 Scout· Stanford HELM specialized suites data docs55.6%
56Qwen1.5-72B-Chat· Stanford HELM specialized suites data docs54.7%
57DeepSeek LLM 67B Chat· Stanford HELM specialized suites data docs54.5%
58Qwen2.5 7B Instruct Turbo· Stanford HELM specialized suites data docs54.1%
59Grok 3· Stanford HELM specialized suites data docs53.8%
60Yi 34B Chat· Stanford HELM specialized suites data docs51.4%
61Grok 3 Mini· Stanford HELM specialized suites data docs49.8%
62Meta Llama 3.1 70B Instruct Turbo· Stanford HELM specialized suites data docs48.2%
63DeepSeek V3· Stanford HELM specialized suites data docs47.3%
64DeepSeek R1· Stanford HELM specialized suites data docs39.2%
65Grok 4· Stanford HELM specialized suites data docs37.4%
66Mixtral 8x22B Instruct v0.1· Stanford HELM specialized suites data docs35.6%
67GLM 4.5 Air· Stanford HELM specialized suites data docs33.3%
68Mistral Large2407· Stanford HELM specialized suites data docs31.5%
69O1 Mini· Stanford HELM specialized suites data docs28.4%
70Mixtral 8x7B Instruct v0.1· Stanford HELM specialized suites data docs22.1%
71Command R+· Stanford HELM specialized suites data docs21.6%
72Mistral Small· Stanford HELM specialized suites data docs20.7%
73Command R Plus· Stanford HELM specialized suites data docs17.3%
74Mistral (7B) Instruct v0.3· Stanford HELM specialized suites data docs7%
75DBRX-instruct· Stanford HELM specialized suites data docs4.7%

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