Provider Voice Arena LeaderboardArtificial Analysis
Compare Text to Speech (TTS) models using each provider's own native voices.
Range | Model | Elo | Samples | Arena Voices | Released | API Pricing 1 | |
|---|---|---|---|---|---|---|---|
| 1 | 1-2 |  Eleven v4 Turbo | 1330±1895% interval 1312 to 1348 | 1,785 | Sept 2026 | $40.0 /1M chars | |
| 2 | 1-2 | Eleven v4 | 1328±1795% interval 1311 to 1345 | 2,078 | Sept 2026 | $80.0 /1M chars | |
| 3 | 3-4 |  Qwen-Audio-3.1-TTS-Plus | 1300±1795% interval 1283 to 1317 | 1,676 | Sept 2026 | $19.3 /1M chars | |
| 4 | 3-5 |  Sonic 3.6 | 1286±1595% interval 1271 to 1301 | 2,462 | Aug 2026 | $49.0 /1M chars | |
| 5 | 4-7 |  Gemini 3.8 Flash TTS | 1274±1595% interval 1259 to 1289 | 2,704 | Sept 2026 | $16.5 /1M chars | |
| 6 | 5-7 | Qwen-Audio-3.0-TTS-Plus | 1268±1595% interval 1253 to 1283 | 2,152 | Sept 2026 | $19.3 /1M chars | |
| 7 | 5-8 |  Realtime TTS-2 | 1260±1595% interval 1245 to 1275 | 1,910 | Aug 2026 | $20.8 /1M chars | |
| 8 | 7-10 | Gemini 3.8 Flash-Lite TTS | 1248±1495% interval 1234 to 1262 | 2,689 | Sept 2026 | $11.0 /1M chars | |
| 9 | 8-10 |  Simba 3.2 | 1245±1395% interval 1232 to 1258 | 3,050 | Jul 2026 | $6.6 /1M chars | |
| 10 | 8-10 |  Fish Audio Drama 3 Preview  | 1243±1695% interval 1227 to 1259 | 1,857 | Oct 2026 | $15.0 /1M chars | |
| 11 | 11-16 |  Breeze TTS 2 | 1221±1495% interval 1207 to 1235 | 2,000 | Aug 2026 | $34.0 /1M chars | |
| 12 | 11-16 |  Luna TTS | 1218±1295% interval 1206 to 1230 | 3,223 | Jun 2026 | $15.0 /1M chars | |
| 13 | 11-16 | Realtime TTS-2 Flash | 1218±1495% interval 1204 to 1232 | 2,338 | Aug 2026 | $10.4 /1M chars | |
| 14 | 11-16 | Gemini 3.1 Flash TTS | 1213±1195% interval 1202 to 1224 | 4,016 | Apr 2026 | $18.3 /1M chars | |
| 15 | 11-16 |  StepAudio 2.5 TTS (Aug 2026) | 1211±1595% interval 1196 to 1226 | 1,892 | Aug 2026 | $85.0 /1M chars | |
| 16 | 11-16 | v3 Conversational | 1209±1495% interval 1195 to 1223 | 2,449 | Aug 2026 | $50.0 /1M chars | |
| 17 | 17-18 | Sonic 3.5 | 1193±1295% interval 1181 to 1205 | 3,525 | May 2026 | $49.0 /1M chars | |
| 18 | 17-19 | Fish Audio S2.1 Pro (Oct 2026) | 1192±1595% interval 1177 to 1207 | 1,780 | Oct 2026 | $15.0 /1M chars | |
| 19 | 19-23 |  Soniox TTS Real-Time v2 | 1180±1195% interval 1169 to 1191 | 5,116 | Aug 2026 | $14.2 /1M chars | |
| 20 | 19-23 | Eleven v3 | 1176±1195% interval 1165 to 1187 | 4,813 | Feb 2026 | $100.0 /1M chars | |
| 21 | 19-23 |  Speech 2.8 HD | 1175±1195% interval 1164 to 1186 | 4,898 | Feb 2026 | $100.0 /1M chars | |
| 22 | 19-23 |  Gradium TTS (Sep 2026) | 1173±1495% interval 1159 to 1187 | 1,878 | Sept 2026 | $47.8 /1M chars | |
| 23 | 19-23 |  Lightning V3.1 Pro (Jul 2026) | 1171±1395% interval 1158 to 1184 | 2,470 | Jul 2026 | $19.5 /1M chars | |
| 24 | 24-27 |  Falcon 2 | 1155±1395% interval 1142 to 1168 | 2,184 | Jul 2026 | $10.0 /1M chars | |
| 25 | 24-27 | Gradium TTS (Aug 2026) | 1152±1395% interval 1139 to 1165 | 2,131 | Aug 2026 | $47.2 /1M chars | |
| 26 | 24-27 | Speech 2.8 Turbo | 1152±1195% interval 1141 to 1163 | 4,724 | Feb 2026 | $60.0 /1M chars | |
| 27 | 24-30 |  SpaceXAI TTS | 1145±1495% interval 1131 to 1159 | 1,586 | Mar 2026 | $15.0 /1M chars | |
| 28 | 27-31 |  Async Flash v1.5 | 1139±1295% interval 1127 to 1151 | 2,430 | May 2026 | $10.1 /1M chars | |
| 29 | 27-31 | Step TTS 2 (Mar 2026) | 1138±1295% interval 1126 to 1150 | 2,560 | Feb 2026 | $40.0 /1M chars | |
| 30 | 27-34 |  Azure HD 2.5 | 1134±1295% interval 1122 to 1146 | 2,743 | Mar 2026 | $22.0 /1M chars | |
| 31 | 28-34 | Speech 2.6 HD | 1130±1195% interval 1119 to 1141 | 4,090 | Oct 2025 | $100.0 /1M chars | |
| 32 | 30-35 | Async Pro v1.0 | 1126±1295% interval 1114 to 1138 | 2,321 | May 2026 | $20.2 /1M chars | |
| 33 | 30-36 | Simba 3.0 | 1123±1295% interval 1111 to 1135 | 2,780 | Feb 2026 | $6.6 /1M chars | |
| 34 | 30-36 | Speech 2.6 Turbo | 1123±1195% interval 1112 to 1134 | 4,405 | Oct 2025 | $60.0 /1M chars | |
| 35 | 32-36 | Fish Audio S2 Pro | 1118±1295% interval 1106 to 1130 | 2,527 | Mar 2026 | $15.0 /1M chars | |
| 36 | 33-38 | Speech-02-HD | 1113±1195% interval 1102 to 1124 | 4,484 | Mar 2025 | $100.0 /1M chars | |
| 37 | 36-42 |  TTS-1 HD | 1106±1195% interval 1095 to 1117 | 3,662 | Nov 2023 | $30.0 /1M chars | |
| 38 | 36-42 |  Chatterbox HD | 1104±1295% interval 1092 to 1116 | 2,517 | — | May 2025 | $40.0 /1M chars |
| 39 | 37-44 | Turbo v2.5 | 1100±1095% interval 1090 to 1110 | 6,923 | Jul 2024 | $50.0 /1M chars | |
| 40 | 37-45 |  Speech 2.6 | 1098±1395% interval 1085 to 1111 | 2,251 | Jul 2026 | N/A | |
| 41 | 37-45 | Step Audio EditX (Mar 2026) | 1097±1295% interval 1085 to 1109 | 2,170 | Feb 2026 | N/A | |
| 42 | 37-45 | Multilingual v2 | 1096±1095% interval 1086 to 1106 | 7,361 | Aug 2023 | $100.0 /1M chars | |
| 43 | 39-48 | Speech-02-Turbo | 1091±1195% interval 1080 to 1102 | 3,859 | Mar 2025 | $60.0 /1M chars | |
| 44 | 39-48 | TTS-1 | 1090±1195% interval 1079 to 1101 | 6,185 | Nov 2023 | $15.0 /1M chars | |
| 45 | 39-50 |  Voxtral TTS | 1088±1295% interval 1076 to 1100 | 2,257 | Mar 2026 | $16.0 /1M chars | |
| 46 | 42-51 | Gemini 2.5 Flash Lite TTS | 1085±1295% interval 1073 to 1097 | 3,201 | Dec 2025 | $9.2 /1M chars | |
| 47 | 43-51 | OpenAudio S1 | 1082±1195% interval 1071 to 1093 | 4,547 | Jun 2025 | $15.0 /1M chars | |
| 48 | 43-51 | Studio | 1081±1195% interval 1070 to 1092 | 4,495 | Feb 2023 | $160.0 /1M chars | |
| 49 | 43-53 | GPT-Realtime-2 | 1079±1395% interval 1066 to 1092 | 2,143 | May 2026 | $191.6 /1M chars | |
| 50 | 45-53 | Flash v2.5 | 1078±1195% interval 1067 to 1089 | 5,738 | Dec 2024 | $50.0 /1M chars | |
| 51 | 46-56 | Journey | 1075±1395% interval 1062 to 1088 | 2,277 | Dec 2023 | $160.0 /1M chars | |
| 52 | 49-60 | T2A-01-HD | 1068±1195% interval 1057 to 1079 | 4,989 | Jan 2025 | $50.0 /1M chars | |
| 53 | 49-61 |  Maya 2 Flash | 1067±1395% interval 1054 to 1080 | 2,012 | Jul 2026 | N/A | |
| 54 | 51-61 |  Polly Generative | 1066±1295% interval 1054 to 1078 | 3,776 | May 2024 | $30.0 /1M chars | |
| 55 | 51-61 |  Magpie-Multilingual 357M (Feb 2026) | 1065±1295% interval 1053 to 1077 | 2,166 | Feb 2026 | N/A | |
| 56 | 52-62 |  Kokoro 82M v1.0 | 1063±1195% interval 1052 to 1074 | 5,357 | Jan 2025 | $0.7 /1M chars | |
| 57 | 52-64 | Chirp 3: HD | 1060±1295% interval 1048 to 1072 | 3,373 | Mar 2025 | $30.0 /1M chars | |
| 58 | 52-64 | Sonic 3 | 1060±1295% interval 1048 to 1072 | 3,510 | Oct 2025 | $49.0 /1M chars | |
| 59 | 52-64 | Gemini 2.5 Flash TTS (Dec 2025) | 1060±1295% interval 1048 to 1072 | 2,898 | Dec 2025 | $9.2 /1M chars | |
| 60 | 52-64 |  Coda | 1059±1295% interval 1047 to 1071 | 2,340 | May 2026 | $50.0 /1M chars | |
| 61 | 52-67 |  MiMo-V2.5-TTS | 1056±1395% interval 1043 to 1069 | 2,015 | - | N/A | |
| 62 | 54-69 |  Flux TTS | 1052±1495% interval 1038 to 1066 | 2,065 | Aug 2026 | $45.0 /1M chars | |
| 63 | 56-68 |  Octave 2 | 1052±1295% interval 1040 to 1064 | 3,121 | Oct 2025 | $87.5 /1M chars | |
| 64 | 57-68 | Async Flash v1.0 | 1051±1195% interval 1040 to 1062 | 4,032 | Jul 2025 | $10.1 /1M chars | |
| 65 | 61-71 | Maya1 | 1047±1295% interval 1035 to 1059 | 3,288 | Jun 2025 | N/A | |
| 66 | 57-73 | Maya 2 Global | 1046±1495% interval 1032 to 1060 | 1,886 | Jul 2026 | N/A | |
| 67 | 56-77 | OpenAudio S1 Mini | 1045±2095% interval 1025 to 1065 | 1,682 | Jun 2025 | $15.0 /1M chars | |
| 68 | 64-73 | Octave TTS | 1041±1195% interval 1030 to 1052 | 4,683 | Feb 2025 | $87.5 /1M chars | |
| 69 | 63-76 |  Higgs Audio V3 TTS | 1039±1395% interval 1026 to 1052 | 2,121 | Jun 2026 | $15.0 /1M chars | |
| 70 | 65-78 |  Bland Speech v3 | 1037±1395% interval 1024 to 1050 | 2,410 | Aug 2026 | $40.0 /1M chars | |
| 71 | 62-78 | Azure Neural | 1036±1995% interval 1017 to 1055 | 1,158 | Sept 2018 | $15.0 /1M chars | |
| 72 | 67-78 | Gemini 2.5 Pro (Dec 2025) | 1033±1395% interval 1020 to 1046 | 2,456 | Dec 2025 | $18.3 /1M chars | |
| 73 | 67-78 | MAI-Voice-1 | 1033±1295% interval 1021 to 1045 | 2,289 | Apr 2026 | N/A | |
| 74 | 68-78 | Polly Long-Form | 1029±1395% interval 1016 to 1042 | 1,955 | Nov 2023 | $100.0 /1M chars | |
| 75 | 69-78 | T2A-01-Turbo | 1029±1195% interval 1018 to 1040 | 4,959 | Jan 2025 | $30.0 /1M chars | |
| 76 | 70-78 | Chatterbox  | 1026±1195% interval 1015 to 1037 | 4,680 | May 2025 | $25.0 /1M chars | |
| 77 | 70-78 | Lightning v3.1 | 1025±1395% interval 1012 to 1038 | 2,241 | Mar 2026 | $25.0 /1M chars | |
| 78 | 70-78 | MiMo-V2-TTS | 1024±1395% interval 1011 to 1037 | 1,834 | Mar 2026 | N/A | |
| 79 | 79-80 |  Raon SpeechLM | 1005±1395% interval 992 to 1018 | 2,138 | Apr 2026 | N/A | |
| 80 | 80 |  Zonos-v0.1 | 1000AnchorFixed at 1000 as the reference point for this board's Elo scale. | 4,855 | Feb 2025 | $20.0 /1M chars | |
| 81 | 81 | Murf Speech Gen 2 | 986±1295% interval 974 to 998 | 4,998 | Jul 2024 | $100.0 /1M chars | |
| 82 | 82-83 | VibeVoice 1.5B | 957±1495% interval 943 to 971 | 2,128 | Aug 2025 | N/A | |
| 83 | 82-83 |  OpenVoice v2 | 956±1395% interval 943 to 969 | 3,020 | — | Apr 2024 | $8.3 /1M chars |
| 84 | 83-87 | Qwen3 TTS Flash | 943±1495% interval 929 to 957 | 2,164 | Sept 2025 | $11.0 /1M chars | |
| 85 | 83-87 |  Neuphonic TTS | 942±1495% interval 928 to 956 | 2,419 | Oct 2024 | $20.8 /1M chars | |
| 86 | 84-87 |  Kugel 3 | 934±1395% interval 921 to 947 | 4,270 | Jun 2026 | N/A | |
| 87 | 84-87 | Qwen3 TTS | 933±1395% interval 920 to 946 | 2,990 | Jan 2026 | $13.8 /1M chars | |
| 88 | 88-90 | WaveNet | 918±1295% interval 906 to 930 | 6,303 | Sept 2016 | $4.0 /1M chars | |
| 89 | 88-90 |  XTTS v2 | 916±1595% interval 901 to 931 | 2,596 | — | Nov 2023 | $40.4 /1M chars |
| 90 | 88-93 | Mist V2 | 907±1595% interval 892 to 922 | 2,929 | Mar 2025 | $31.1 /1M chars | |
| 91 | 90-94 | Polly Neural | 896±1495% interval 882 to 910 | 2,878 | Jul 2019 | $16.0 /1M chars | |
| 92 | 90-94 | Neural2 | 895±1295% interval 883 to 907 | 6,583 | Jun 2022 | $16.0 /1M chars | |
| 93 | 90-94 |  StyleTTS 2 | 895±1595% interval 880 to 910 | 2,560 | — | Jun 2023 | $2.8 /1M chars |
| 94 | 91-95 | Standard | 887±1295% interval 875 to 899 | 6,642 | Aug 2018 | $4.0 /1M chars | |
| 95 | 94-95 |  Noiz TTS | 878±1595% interval 863 to 893 | 2,238 | Nov 2025 | N/A | |
| 96 | 96-97 |  MetaVoice v1 | 848±1895% interval 830 to 866 | 1,552 | — | Feb 2024 | N/A |
| 97 | 97 | Polly Standard | 832±1695% interval 816 to 848 | 3,494 | Nov 2016 | $4.0 /1M chars |
1 API Pricing reflects the cost to generate 1M characters on the model creator's API with the model's default settings
2 Open Weights includes models that can be self-hosted using downloadable model weights or a provider-supplied inference runtime. Open Weights Commercial includes models whose published license permits commercial use, including use subject to license conditions (e.g. an NVIDIA AI Enterprise subscription), and excludes models that require a separately negotiated commercial agreement.
Frequently Asked Questions
The top Text to Speech models by Elo rating are: 1. Eleven v4 Turbo (Elo 1330), 2. Eleven v4 (Elo 1328), 3. Qwen-Audio-3.1-TTS-Plus (Elo 1300), 4. Sonic 3.6 (Elo 1286), and 5. Gemini 3.8 Flash TTS (Elo 1274). Rankings are based on blind user votes in the Speech Arena.
Kokoro 82M v1.0 is the most affordable at $0.65 per 1M characters with an Elo score of 1063. Other affordable options include StyleTTS 2 at $2.82 per 1M characters.
The top 5 open weights Text to Speech models are: 1. Fish Audio Drama 3 Preview (Elo 1243), 2. Breeze TTS 2 (Elo 1221), 3. Fish Audio S2 Pro (Elo 1118), 4. Step Audio EditX (Mar 2026) (Elo 1097), and 5. Voxtral TTS (Elo 1088).