Artificial Analysis Speech Arena Leaderboard
Quality evaluation of Text to Speech Models based on the Speech Arena of crowdsourced preferences.
To participate in the Speech Arena see the Speech Arena tab.
For further details on methodology and analysis of speed & price, see artificialanalysis.ai.
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 | Pricing 1 | |
|---|---|---|---|---|---|---|---|
| 1 | 1-2 | Eleven v4 Turbo | 1340±1595% interval 1325 to 1355 | 2,559 | Sept 2026 | $40.00 /1M chars | |
| 2 | 1-2 | Eleven v4 | 1338±1595% interval 1323 to 1353 | 2,853 | Sept 2026 | $80.00 /1M chars | |
| 3 | 3-5 | Qwen-Audio-3.1-TTS-Plus | 1298±1495% interval 1284 to 1312 | 2,614 | Sept 2026 | $19.31 /1M chars | |
| 4 | 3-6 | Sonic 3.6 | 1295±1395% interval 1282 to 1308 | 3,499 | Aug 2026 | $49.00 /1M chars | |
| 5 | 3-7 | Simba 3.5 | 1291±1795% interval 1274 to 1308 | 1,852 | Nov 2026 | $6.58 /1M chars | |
| 6 | 4-8 | Gemini 3.8 Flash TTS | 1283±1395% interval 1270 to 1296 | 3,772 | Sept 2026 | $16.49 /1M chars | |
| 7 | 4-10 | 1279±1795% interval 1262 to 1296 | 1,861 | Oct 2026 | $15.00 /1M chars | ||
| 8 | 6-11 | Lightning V3.1 Pro (Oct 2026) | 1271±1595% interval 1256 to 1286 | 2,173 | Oct 2026 | $19.50 /1M chars | |
| 9 | 7-11 | Realtime TTS-2 | 1270±1395% interval 1257 to 1283 | 2,884 | Aug 2026 | $20.83 /1M chars | |
| 10 | 8-12 | Qwen-Audio-3.0-TTS-Plus | 1263±1395% interval 1250 to 1276 | 3,173 | Sept 2026 | $19.31 /1M chars | |
| 11 | 8-12 | Gemini 3.8 Flash-Lite TTS | 1259±1295% interval 1247 to 1271 | 3,741 | Sept 2026 | $11.03 /1M chars | |
| 12 | 11-12 | Simba 3.2 | 1250±1195% interval 1239 to 1261 | 4,044 | Jul 2026 | $6.58 /1M chars | |
| 13 | 13-14 | 1236±1395% interval 1223 to 1249 | 2,882 | Oct 2026 | $15.00 /1M chars | ||
| 14 | 13-14 | 1229±1295% interval 1217 to 1241 | 3,006 | Aug 2026 | $34.00 /1M chars | ||
| 15 | 15-20 | v3 Conversational | 1216±1295% interval 1204 to 1228 | 3,445 | Aug 2026 | $50.00 /1M chars | |
| 16 | 15-20 | StepAudio 2.5 TTS (Aug 2026) | 1215±1295% interval 1203 to 1227 | 2,893 | Aug 2026 | $85.00 /1M chars | |
| 17 | 15-20 | Gemini 3.1 Flash TTS | 1215±1095% interval 1205 to 1225 | 4,879 | Apr 2026 | $18.31 /1M chars | |
| 18 | 15-20 | Realtime TTS-2 Flash | 1215±1295% interval 1203 to 1227 | 3,323 | Aug 2026 | $10.42 /1M chars | |
| 19 | 15-20 | Voxtral 3 TTS | 1213±1395% interval 1200 to 1226 | 2,340 | Oct 2026 | $13.00 /1M chars | |
| 20 | 15-20 | AssemblyAI-TTS-1 | 1212±1595% interval 1197 to 1227 | 1,947 | Oct 2026 | $15.00 /1M chars | |
| 21 | 21-23 | Luna TTS | 1196±1195% interval 1185 to 1207 | 4,120 | Jun 2026 | $15.00 /1M chars | |
| 22 | 21-23 | Sonic 3.5 | 1195±1095% interval 1185 to 1205 | 4,387 | May 2026 | $49.00 /1M chars | |
| 23 | 21-26 | Eleven v3 | 1190±1095% interval 1180 to 1200 | 5,607 | Feb 2026 | $100.00 /1M chars | |
| 24 | 23-29 | 1182±1395% interval 1169 to 1195 | 2,712 | Oct 2026 | N/A | ||
| 25 | 23-28 | Fish Audio S2.1 Pro (Oct 2026) | 1181±1295% interval 1169 to 1193 | 2,711 | Oct 2026 | $15.00 /1M chars | |
| 26 | 23-28 | Soniox TTS Real-Time v2 | 1181±1095% interval 1171 to 1191 | 5,999 | Aug 2026 | $14.23 /1M chars | |
| 27 | 24-29 | Speech 2.8 HD | 1174±1095% interval 1164 to 1184 | 5,705 | Feb 2026 | $100.00 /1M chars | |
| 28 | 24-29 | Gradium TTS (Sep 2026) | 1173±1295% interval 1161 to 1185 | 2,818 | Sept 2026 | $47.78 /1M chars | |
| 29 | 24-31 | Gradium TTS (Oct 2026) | 1169±1395% interval 1156 to 1182 | 2,225 | Sept 2026 | $47.78 /1M chars | |
| 30 | 29-35 | SpaceXAI TTS | 1160±1295% interval 1148 to 1172 | 2,357 | Mar 2026 | $15.00 /1M chars | |
| 31 | 29-35 | Falcon 2 | 1159±1295% interval 1147 to 1171 | 2,973 | Jul 2026 | $10.00 /1M chars | |
| 32 | 30-35 | Gradium TTS (Aug 2026) | 1155±1295% interval 1143 to 1167 | 3,029 | Aug 2026 | $47.78 /1M chars | |
| 33 | 30-36 | Falcon 2.1 | 1153±1595% interval 1138 to 1168 | 1,514 | Oct 2026 | $10.00 /1M chars | |
| 34 | 30-35 | Speech 2.8 Turbo | 1151±1095% interval 1141 to 1161 | 5,429 | Feb 2026 | $60.00 /1M chars | |
| 35 | 30-36 | Azure HD 2.5 | 1150±1195% interval 1139 to 1161 | 3,383 | Mar 2026 | $22.00 /1M chars | |
| 36 | 35-38 | Async Flash v1.5 | 1139±1195% interval 1128 to 1150 | 3,126 | May 2026 | $10.10 /1M chars | |
| 37 | 36-38 | Step TTS 2 (Mar 2026) | 1136±1195% interval 1125 to 1147 | 3,190 | Feb 2026 | $40.00 /1M chars | |
| 38 | 36-40 | Async Pro v1.0 | 1130±1195% interval 1119 to 1141 | 2,989 | May 2026 | $20.20 /1M chars | |
| 39 | 38-41 | Speech 2.6 HD | 1124±1095% interval 1114 to 1134 | 4,648 | Oct 2025 | $100.00 /1M chars | |
| 40 | 38-42 | Speech 2.6 Turbo | 1123±1095% interval 1113 to 1133 | 4,882 | Oct 2025 | $60.00 /1M chars | |
| 41 | 39-48 | Coda v2 | 1114±1395% interval 1101 to 1127 | 2,009 | Oct 2026 | $50.00 /1M chars | |
| 42 | 39-46 | Simba 3.0 | 1113±1195% interval 1102 to 1124 | 3,391 | Feb 2026 | $6.58 /1M chars | |
| 43 | 41-48 | Chatterbox HD | 1110±1195% interval 1099 to 1121 | 3,003 | — | May 2025 | $40.00 /1M chars |
| 44 | 41-48 | 1109±1395% interval 1096 to 1122 | 2,000 | Oct 2026 | $0.00 /1M chars | ||
| 45 | 41-48 | TTS-1 HD | 1108±1195% interval 1097 to 1119 | 4,069 | Nov 2023 | $30.00 /1M chars | |
| 46 | 41-51 | Speech-02-HD | 1103±1195% interval 1092 to 1114 | 4,939 | Mar 2025 | $100.00 /1M chars | |
| 47 | 43-52 | 1102±1195% interval 1091 to 1113 | 2,649 | Feb 2026 | N/A | ||
| 48 | 43-53 | 1102±1195% interval 1091 to 1113 | 3,042 | Mar 2026 | $15.00 /1M chars | ||
| 49 | 46-56 | 1094±1195% interval 1083 to 1105 | 2,787 | Mar 2026 | $16.00 /1M chars | ||
| 50 | 46-54 | Turbo v2.5 | 1094±1095% interval 1084 to 1104 | 7,312 | Jul 2024 | $50.00 /1M chars | |
| 51 | 47-56 | Multilingual v2 | 1093±1095% interval 1083 to 1103 | 7,722 | Aug 2023 | $100.00 /1M chars | |
| 52 | 47-56 | Speech 2.6 | 1092±1195% interval 1081 to 1103 | 2,880 | Jul 2026 | N/A | |
| 53 | 48-56 | Speech-02-Turbo | 1091±1195% interval 1080 to 1102 | 4,298 | Mar 2025 | $60.00 /1M chars | |
| 54 | 49-56 | OpenAudio S1 | 1090±1195% interval 1079 to 1101 | 4,885 | Jun 2025 | $15.00 /1M chars | |
| 55 | 51-57 | TTS-1 | 1084±1095% interval 1074 to 1094 | 6,549 | Nov 2023 | $15.00 /1M chars | |
| 56 | 49-59 | Gemini 2.5 Flash Lite TTS | 1083±1195% interval 1072 to 1094 | 3,545 | Dec 2025 | $9.15 /1M chars | |
| 57 | 55-62 | Flash v2.5 | 1075±1095% interval 1065 to 1085 | 6,063 | Dec 2024 | $50.00 /1M chars | |
| 58 | 55-64 | T2A-01-HD | 1074±1195% interval 1063 to 1085 | 5,290 | Jan 2025 | $50.00 /1M chars | |
| 59 | 55-64 | Studio | 1074±1195% interval 1063 to 1085 | 4,827 | Feb 2023 | $160.00 /1M chars | |
| 60 | 56-68 | GPT-Realtime-2 | 1072±1295% interval 1060 to 1084 | 2,609 | May 2026 | $191.56 /1M chars | |
| 61 | 57-68 | Gemini 2.5 Flash TTS (Dec 2025) | 1071±1295% interval 1059 to 1083 | 3,167 | Dec 2025 | $9.15 /1M chars | |
| 62 | 57-69 | Journey | 1067±1295% interval 1055 to 1079 | 2,563 | Dec 2023 | $160.00 /1M chars | |
| 63 | 57-70 | 1063±1295% interval 1051 to 1075 | 2,564 | Feb 2026 | N/A | ||
| 64 | 58-69 | Polly Generative | 1063±1195% interval 1052 to 1074 | 4,030 | May 2024 | $30.00 /1M chars | |
| 65 | 60-71 | Chirp 3: HD | 1061±1195% interval 1050 to 1072 | 3,662 | Mar 2025 | $30.00 /1M chars | |
| 66 | 61-71 | 1060±1195% interval 1049 to 1071 | 5,662 | Jan 2025 | $0.65 /1M chars | ||
| 67 | 61-72 | Sonic 3 | 1060±1195% interval 1049 to 1071 | 3,785 | Oct 2025 | $49.00 /1M chars | |
| 68 | 60-72 | Maya 2 Flash | 1060±1295% interval 1048 to 1072 | 2,507 | Jul 2026 | N/A | |
| 69 | 62-74 | Octave 2 | 1058±1295% interval 1046 to 1070 | 3,422 | Oct 2025 | $87.50 /1M chars | |
| 70 | 63-76 | Flux TTS | 1052±1295% interval 1040 to 1064 | 2,541 | Aug 2026 | $45.00 /1M chars | |
| 71 | 65-77 | Coda | 1051±1295% interval 1039 to 1063 | 2,762 | May 2026 | $50.00 /1M chars | |
| 72 | 66-78 | MiMo-V2.5-TTS | 1049±1295% interval 1037 to 1061 | 2,388 | - | N/A | |
| 73 | 69-79 | 1048±1295% interval 1036 to 1060 | 3,552 | Jun 2025 | N/A | ||
| 74 | 68-79 | Maya 2 Global | 1047±1395% interval 1034 to 1060 | 2,295 | Jul 2026 | N/A | |
| 75 | 70-79 | Async Flash v1.0 | 1046±1195% interval 1035 to 1057 | 4,282 | Jul 2025 | $10.10 /1M chars | |
| 76 | 65-84 | 1043±2095% interval 1023 to 1063 | 1,678 | Jun 2025 | $15.00 /1M chars | ||
| 77 | 72-81 | Octave TTS | 1039±1195% interval 1028 to 1050 | 4,889 | Feb 2025 | $87.50 /1M chars | |
| 78 | 72-83 | Gemini 2.5 Pro (Dec 2025) | 1038±1295% interval 1026 to 1050 | 2,668 | Dec 2025 | $18.31 /1M chars | |
| 79 | 73-84 | 1036±1295% interval 1024 to 1048 | 2,454 | Jun 2026 | $15.00 /1M chars | ||
| 80 | 76-85 | Bland Speech v3 | 1033±1295% interval 1021 to 1045 | 2,888 | Aug 2026 | $40.00 /1M chars | |
| 81 | 70-86 | Azure Neural | 1033±2095% interval 1013 to 1053 | 1,156 | Sept 2018 | $15.00 /1M chars | |
| 82 | 77-86 | Polly Long-Form | 1027±1395% interval 1014 to 1040 | 2,160 | Nov 2023 | $100.00 /1M chars | |
| 83 | 78-86 | MAI-Voice-1 | 1026±1295% interval 1014 to 1038 | 2,530 | Apr 2026 | N/A | |
| 84 | 80-86 | T2A-01-Turbo | 1025±1195% interval 1014 to 1036 | 5,173 | Jan 2025 | $30.00 /1M chars | |
| 85 | 80-86 | Lightning v3.1 | 1022±1295% interval 1010 to 1034 | 2,602 | Mar 2026 | $25.00 /1M chars | |
| 86 | 82-86 | 1020±1195% interval 1009 to 1031 | 4,890 | May 2025 | $25.00 /1M chars | ||
| 87 | 87-88 | Raon SpeechLM | 1007±1395% interval 994 to 1020 | 2,495 | Apr 2026 | N/A | |
| 88 | 88 | 1000AnchorFixed at 1000 as the reference point for this board's Elo scale. | 5,023 | Feb 2025 | $20.00 /1M chars | ||
| 89 | 89 | Murf Speech Gen 2 | 986±1295% interval 974 to 998 | 5,151 | Jul 2024 | $100.00 /1M chars | |
| 90 | 90-91 | 955±1395% interval 942 to 968 | 3,181 | — | Apr 2024 | $8.33 /1M chars | |
| 91 | 90-91 | 955±1495% interval 941 to 969 | 2,249 | Aug 2025 | N/A | ||
| 92 | 92-94 | Qwen3 TTS Flash | 940±1495% interval 926 to 954 | 2,293 | Sept 2025 | $11.03 /1M chars | |
| 93 | 92-95 | Neuphonic TTS | 939±1495% interval 925 to 953 | 2,567 | Oct 2024 | $20.79 /1M chars | |
| 94 | 92-95 | Qwen3 TTS | 929±1395% interval 916 to 942 | 3,212 | Jan 2026 | $13.79 /1M chars | |
| 95 | 94-97 | Kugel 3 | 926±1395% interval 913 to 939 | 4,630 | Jun 2026 | N/A | |
| 96 | 95-97 | WaveNet | 915±1295% interval 903 to 927 | 6,469 | Sept 2016 | $4.00 /1M chars | |
| 97 | 94-98 | 915±1595% interval 900 to 930 | 2,766 | — | Nov 2023 | $40.44 /1M chars | |
| 98 | 96-101 | Mist V2 | 903±1595% interval 888 to 918 | 3,132 | Mar 2025 | $31.13 /1M chars | |
| 99 | 98-102 | Polly Neural | 894±1495% interval 880 to 908 | 3,027 | Jul 2019 | $16.00 /1M chars | |
| 100 | 98-102 | Neural2 | 892±1295% interval 880 to 904 | 6,743 | Jun 2022 | $16.00 /1M chars | |
| 101 | 98-102 | 888±1595% interval 873 to 903 | 2,761 | — | Jun 2023 | $2.82 /1M chars | |
| 102 | 99-102 | Standard | 884±1395% interval 871 to 897 | 6,819 | Aug 2018 | $4.00 /1M chars | |
| 103 | 103 | Noiz TTS | 867±1495% interval 853 to 881 | 2,555 | Nov 2025 | N/A | |
| 104 | 104-105 | 845±1895% interval 827 to 863 | 1,712 | — | Feb 2024 | N/A | |
| 105 | 104-105 | Polly Standard | 830±1695% interval 814 to 846 | 3,642 | Nov 2016 | $4.00 /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.