WoodstitchHuman evals for voice AI

TTS voice leaderboard · US English · preliminary

Best TTS Voices for Nonfiction Narration

Stock text-to-speech voices ranked by blind paired listening on nonfiction book passages.

Voices53 from 19 providersJudgments1,228 so farListeners74 paid, USUpdated21 Nov 2026, 21:00 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

#RangeVoiceScoreEloJudgmentsPlay
11–39 KoreGoogle Gemini TTS
1117±19746
21–32 KelseySmallest.ai
1115±18251
31–41 JoleneCartesia
1114±20148
41–41 CelesteRime
1114±20337
51–41 Spuds OxleyElevenLabs
1111±19751
61–39 af_heartKokoro
1106±22432
71–43 MarinOpenAI
1105±19242
81–39 KatieCartesia
1100±18552
91–39 DanielleAmazon Polly
1089±21147
101–42 am_michaelKokoro
1082±22836
112–43 HenrySpeechify
1062±17050
122–43 GeorgeSpeechify
1061±14956
131–49 ClementineRime
1055±21545
141–45 JonElevenLabs
1052±17651
152–48 AdrianFish Audio
1048±19944
162–45 CharonGoogle Cloud TTS
1046±16050
174–44 OrusGoogle Cloud TTS
1037±13964
182–51 DefaultChatterbox
1031±21027
192–43 AshleyInworld
1031±16656
204–46 GeffenSpeechify
1030±14450
213–49 LaurenCartesia
1025±18755
223–49 CaptivatingStorytellerMiniMax
1011±18859
232–49 EldrinInworld
1009±19654
243–51 Expressive narratorMiniMax
1004±21249
255–48 ChelseaSmallest.ai
1003±18055
265–50 CoralOpenAI
1002±17551
273–48 AlgenibGoogle Gemini TTS
1002±18151
285–50 AlaricInworld
996±16654
294–49 TaraOrpheus
995±17933
305–49 Wayne (M)Murf
995±18252
3111–49 ThaliaDeepgram
990±13752
326–50 PeterElevenLabs
985±17245
334–51 CedarOpenAI
982±20643
348–50 Daniel (M)Murf
982±16048
352–51 af_bellaKokoro
974±20434
366–50 AchernarGoogle Gemini TTS
974±17753
378–51 AthenaDeepgram
972±16650
386–50 HannahFish Audio
969±17642
397–52 SavannahZonos3-Mini
968±18437
409–50 AoedeGoogle Cloud TTS
966±17455
419–52 SarahFish Audio
964±17050
426–52 PatrickAmazon Polly
950±20348
4312–52 LyraRime
945±15550
445–53 RyanQwen3-TTS
943±23527
4512–51 LeoSmallest.ai
936±15357
4613–52 PatientManMiniMax
923±18352
4713–53 NatalieMurf
918±20652
4813–53 RuthAmazon Polly
894±20543
494–53 WyattZonos3-Mini
887±30725
5030–53 CoraDeepgram
847±15549
5119–53 AidenQwen3-TTS
840±25834
5221–53 LeoOrpheus
829±26334
5315–53 ZoeOrpheus
812±29128

Listener question: “Which narration would you choose to continue listening to for this nonfiction book?” · 1,228 judgments from 74 listeners so far · 53 voices ranked; the rest are still collecting. Range = ranks consistent with the data at 95%. Methodology

Independent blind listening evaluation. Paid, vetted US listeners.
Methodology · Evaluated audio · Judgment data · Contact