Skip to content

Benchmark

Artificial Analysis Full Duplex Bench · Conversational Dynamics

Scores are percentages. Higher is better.

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

Score distribution

97%
Mean
97%
Median
0%
Std dev
97%
10th percentile
97%
90th percentile
97%
Min

Leaderboard

#ModelScore
1Grok Voice 3.0October 13 release table · Artificial Analysis Full Duplex Bench subset· Reported by SpaceXAI, Artificial Analysis· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.97%

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.