Skip to content

Benchmark

Artificial Analysis · Input Audio Cost on Big Bench Audio Subset

Lower scores are better.

Models measured
3
Scores recorded
3
Sources
1
IQ weight
No weight published

Score distribution

3
Mean
3.5
Median
1.6
Std dev
0.84
10th percentile
4.8
90th percentile
0.84
Min
0.844.8
The band covers the middle 80% of scores. The line marks the median.

Results

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

ModelScore
Gemini 3.8 LiveGemini API · Big Bench Audio subset · thinking level not printed on chart· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.0.84
Gemini 3.8 Live Extended Thinkinghigh thinking · Gemini API · Big Bench Audio subset· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.3.5
Grok Voice 3.0Artificial Analysis live leaderboard · October 24 snapshot · 40-question Big Bench Audio subset· Reported by Artificial Analysis· via BenchBoard docs4.8

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.