TTS voice leaderboard · US English · preliminary
Best TTS Voices for Phone Support
Stock text-to-speech voices ranked by blind paired listening on support-agent lines, heard over a phone connection.
Voices53 from 19 providersJudgments543 so farListeners37 paid, USUpdated20 Nov 2026, 20:30 UTCHow we evaluate →
Top 20 voices
Elo-scale strength from blind paired listening (1000 = average voice in this use case) · whiskers = 95% interval · higher is better
| # | Range | Voice | Score | Elo | Judgments | Play |
|---|---|---|---|---|---|---|
| 1 | 1–23 | 1197±253 | 29 | |||
| 2 | 1–35 | 1166±348 | 22 | |||
| 3 | 1–31 | 1159±291 | 23 | |||
| 4 | 1–33 | 1159±360 | 22 | |||
| 5 | 1–38 | 1145±418 | 18 | |||
| 6 | 1–35 | 1144±305 | 28 | |||
| 7 | 2–31 | 1137±235 | 29 | |||
| 8 | 1–35 | 1125±286 | 21 | |||
| 9 | 1–35 | 1122±315 | 27 | |||
| 10 | 1–36 | 1116±289 | 23 | |||
| 11 | 1–35 | 1105±300 | 28 | |||
| 12 | 1–35 | 1104±282 | 22 | |||
| 13 | 3–34 | 1094±200 | 30 | |||
| 14 | 1–35 | 1087±288 | 22 | |||
| 15 | 3–36 | 1077±253 | 26 | |||
| 16 | 2–40 | 1075±379 | 21 | |||
| 17 | 2–33 | 1074±250 | 26 | |||
| 18 | 1–38 | 1071±299 | 23 | |||
| 19 | 3–37 | 1060±257 | 23 | |||
| 20 | 3–38 | 1054±249 | 21 | |||
| 21 | 3–37 | 1041±245 | 28 | |||
| 22 | 3–40 | 1030±320 | 19 | |||
| 23 | 1–39 | 1023±319 | 23 | |||
| 24 | 2–38 | 1019±265 | 25 | |||
| 25 | 8–40 | 1000±251 | 22 | |||
| 26 | 3–40 | 995±288 | 28 | |||
| 27 | 5–39 | 993±231 | 25 | |||
| 28 | 5–40 | 991±275 | 27 | |||
| 29 | 2–39 | 990±311 | 29 | |||
| 30 | 5–41 | 974±301 | 22 | |||
| 31 | 5–41 | 969±307 | 17 | |||
| 32 | 10–40 | 949±233 | 28 | |||
| 33 | 9–41 | 934±256 | 26 | |||
| 34 | 11–40 | 933±230 | 29 | |||
| 35 | 15–40 | 929±200 | 25 | |||
| 36 | 10–41 | 911±274 | 29 | |||
| 37 | 10–41 | 907±266 | 32 | |||
| 38 | 9–42 | 904±324 | 24 | |||
| 39 | 16–42 | 861±293 | 25 | |||
| 40 | 15–42 | 849±316 | 25 | |||
| 41 | 28–42 | 652±297 | 24 | |||
| 42 | 32–42 | 593±274 | 26 | |||
| — | — | collecting (8 so far) | 8 | |||
| — | — | collecting (6 so far) | 6 | |||
| — | — | collecting (5 so far) | 5 | |||
| — | — | collecting (5 so far) | 5 | |||
| — | — | collecting (5 so far) | 5 | |||
| — | — | collecting (4 so far) | 4 | |||
| — | — | collecting (4 so far) | 4 | |||
| — | — | collecting (3 so far) | 3 | |||
| — | — | collecting (2 so far) | 2 | |||
| — | — | collecting (2 so far) | 2 | |||
| — | — | collecting (0 so far) | 0 | |||
Listener question: “If you needed help with this issue, which voice would you prefer to hear from the support agent?” · 543 judgments from 37 listeners so far · 42 voices ranked; the rest are still collecting. Range = ranks consistent with the data at 95%. Phone clips pass through a narrowband phone condition; this is a listening test, not an agent benchmark. Methodology