Skip to content

Benchmark

AHELM: MuTox - EM

Scores are percentages.

Models measured
12
Scores recorded
359
Sources
1
IQ weight
No weight published

Score distribution

75.5%
Mean
77.9%
Median
15.6%
Std dev
53.4%
10th percentile
100%
90th percentile
33.3%
Min
33.3%100%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
101gemini-1.5-pro-001language: Turkish· AHELM (HELM Audio) data85.7%
102gemini-1.5-pro-002language: Turkish· AHELM (HELM Audio) data85.7%
103Qwen2.5-Omni 7Blanguage: German· AHELM (HELM Audio) data85.7%
104Qwen2.5-Omni 7Blanguage: Hindi· AHELM (HELM Audio) data85.7%
105Qwen2.5-Omni 7Blanguage: Turkish· AHELM (HELM Audio) data85.7%
106GPT-4o mini Audio Previewlanguage: Czech· AHELM (HELM Audio) data85%
107Gemini 2.0 Flash Litelanguage: Catalan· AHELM (HELM Audio) data84.9%
108gemini-1.5-pro-001language: Catalan· AHELM (HELM Audio) data84.9%
109Qwen2.5-Omni 7Blanguage: Hebrew· AHELM (HELM Audio) data84.7%
110Gemini 2.5 Pro Preview 05-06language: Finnish· AHELM (HELM Audio) data84.7%
111Gemini 2.0 Flashlanguage: Dutch· AHELM (HELM Audio) data84.6%
112Gemini 2.0 Flash Experimentallanguage: Dutch· AHELM (HELM Audio) data84.6%
113Qwen2.5-Omni 7Blanguage: Polish· AHELM (HELM Audio) data84.6%
114Gemini 2.5 Pro Preview 05-06language: Bulgarian· AHELM (HELM Audio) data84.4%
115GPT-4o Audio Previewlanguage: Bulgarian· AHELM (HELM Audio) data84.4%
116Gemini 2.5 Flash Preview 05-20 (no think)language: Czech· AHELM (HELM Audio) data84.1%
117Qwen2.5-Omni 7Blanguage: Czech· AHELM (HELM Audio) data84.1%
118GPT-4o mini Audio Previewlanguage: Danish· AHELM (HELM Audio) data83.6%
119GPT-4o mini Audio Previewlanguage: Bulgarian· AHELM (HELM Audio) data83.4%
120Qwen2.5-Omni 7Blanguage: Estonian· AHELM (HELM Audio) data83.1%
121GPT-4o Audio Previewlanguage: Hungarian· AHELM (HELM Audio) data83.1%
122Qwen2.5-Omni 7Blanguage: Bulgarian· AHELM (HELM Audio) data82.9%
123Qwen2.5-Omni 7B· AHELM (HELM Audio) data82.8%
124Gemini 2.0 Flash Litelanguage: Greek· AHELM (HELM Audio) data82.4%
125Gemini 2.0 Flash Litelanguage: Bengali· AHELM (HELM Audio) data82.4%
126Gemini 2.5 Flash Preview 05-20 (no think)language: Bengali· AHELM (HELM Audio) data82.4%
127gemini-1.5-flash-001language: Bengali· AHELM (HELM Audio) data82.4%
128gemini-1.5-pro-001language: Bengali· AHELM (HELM Audio) data82.4%
129Gemini 2.0 Flash Litelanguage: Czech· AHELM (HELM Audio) data82.3%
130Gemini 2.5 Flash Preview 05-20 (no think)language: Slovak· AHELM (HELM Audio) data82.1%
131gemini-1.5-pro-002· AHELM (HELM Audio) data81.9%
132GPT-4o Audio Previewlanguage: Danish· AHELM (HELM Audio) data81.9%
133Gemini 2.5 Pro Preview 05-06language: Tagalog· AHELM (HELM Audio) data81.8%
134gemini-1.5-flash-001language: Tagalog· AHELM (HELM Audio) data81.8%
135gemini-1.5-pro-002language: Slovak· AHELM (HELM Audio) data81.5%
136GPT-4o mini Audio Previewlanguage: Italian· AHELM (HELM Audio) data81.3%
137Qwen2.5-Omni 7Blanguage: Italian· AHELM (HELM Audio) data81.3%
138Gemini 2.0 Flash Lite· AHELM (HELM Audio) data81.2%
139Gemini 2.0 Flash Litelanguage: Polish· AHELM (HELM Audio) data81.1%
140gemini-1.5-pro-002language: Finnish· AHELM (HELM Audio) data81%
141GPT-4o mini Audio Previewlanguage: Hungarian· AHELM (HELM Audio) data80.5%
142Gemini 2.0 Flash Litelanguage: Slovak· AHELM (HELM Audio) data80.3%
143Gemini 2.0 Flashlanguage: Swahili· AHELM (HELM Audio) data80%
144Gemini 2.0 Flash Experimentallanguage: Swahili· AHELM (HELM Audio) data80%
145Gemini 2.0 Flash Litelanguage: Swahili· AHELM (HELM Audio) data80%
146Gemini 2.5 Flash Preview 05-20 (no think)language: Arabic· AHELM (HELM Audio) data80%
147Gemini 2.5 Flash Preview 05-20 (no think)language: Swahili· AHELM (HELM Audio) data80%
148Gemini 2.5 Pro Preview 05-06language: Arabic· AHELM (HELM Audio) data80%
149gemini-1.5-flash-001language: Indonesian· AHELM (HELM Audio) data80%
150gemini-1.5-pro-001language: Swahili· AHELM (HELM Audio) data80%

Showing 101–150 of 359 results.