AA Speech to Speech - Tau-Voice 2.0 - Retail: leaderboard

Sierra tau-bench v2.0.0 voice: 50 English tasks each in airline, retail and telecom; three trials; GPT5.5 low simulated caller speaking ElevenLabs eleven_v3; common prompt/tools/policy, 2400-second call limit. State/action/output checks; infrastructure errors excluded and rerun. Overall is equal domain mean. GPT-Live's source-declared backend prompt differs. Separate from V1; no cross-edition score replacement. Metric: Pass@1 (%), mean of three trials. Source: https://artificialanalysis.ai/speech-to-speech

Metric: Pass@1 (%), mean of three trials. Source: artificialanalysis.ai. Saturation forecast: Rough model projection: around 2026. 11 models tracked. Updated 17 Nov 2026.

Top models

#ModelSettingScoreOverall rank
1Phonic 1High78
2GPT-Live 1 (Sol, low)72.67
3GPT-Live 1 (Astra, medium)71.33
4Qwen Audio 3.0 Realtime Plus55.33
5Grok Voice 3.054
6Gemini 3.8 Live (Extended Thinking)High50
7Eleven Realtime v1 Mini41.33
8Eleven Realtime v1 Max39.33
9Gemini 3.8 Live34
10GPT Realtime 2.1High31.33
11Gemini 3.1 Flash Live High16

Interactive version: theaggregate.ai/benchmark?slug=aa-speech-to-speech-tau-voice-2-0-retail · How It Works · Data refreshed daily, snapshot 2026-11-17.