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 |  Eleven v4 | 1320±1895% interval 1302 to 1338 | 1,803 | Sept 2026 | $80.0 /1M chars | |
| 2 | 2-4 |  Sonic 3.6 | 1273±1695% interval 1257 to 1289 | 2,093 | Aug 2026 | $49.0 /1M chars | |
| 3 | 2-4 |  Gemini 3.8 Flash TTS | 1272±1695% interval 1256 to 1288 | 2,330 | Sept 2026 | $16.5 /1M chars | |
| 4 | 2-5 |  Qwen-Audio-3.0-TTS-Plus | 1259±1695% interval 1243 to 1275 | 1,747 | Sept 2026 | $19.3 /1M chars | |
| 5 | 4-7 |  Realtime TTS-2 | 1247±1795% interval 1230 to 1264 | 1,513 | Aug 2026 | $20.8 /1M chars | |
| 6 | 5-8 | Gemini 3.8 Flash-Lite TTS | 1240±1595% interval 1225 to 1255 | 2,294 | Sept 2026 | $11.0 /1M chars | |
| 7 | 5-8 |  Simba 3.2 | 1239±1395% interval 1226 to 1252 | 2,669 | Jul 2026 | $6.6 /1M chars | |
| 8 | 7-8 |  Luna TTS | 1226±1395% interval 1213 to 1239 | 2,804 | Jun 2026 | $15.0 /1M chars | |
| 9 | 9-13 | Realtime TTS-2 Flash | 1209±1595% interval 1194 to 1224 | 1,943 | Aug 2026 | $10.4 /1M chars | |
| 10 | 9-13 |  Breeze TTS 2  | 1206±1695% interval 1190 to 1222 | 1,575 | Aug 2026 | $34.0 /1M chars | |
| 11 | 9-13 |  StepAudio 2.5 TTS (Aug 2026) | 1206±1695% interval 1190 to 1222 | 1,494 | Aug 2026 | $85.0 /1M chars | |
| 12 | 9-13 | Gemini 3.1 Flash TTS | 1205±1295% interval 1193 to 1217 | 3,622 | Apr 2026 | $18.3 /1M chars | |
| 13 | 9-14 | v3 Conversational | 1200±1595% interval 1185 to 1215 | 2,039 | Aug 2026 | $50.0 /1M chars | |
| 14 | 14 | Sonic 3.5 | 1187±1295% interval 1175 to 1199 | 3,128 | May 2026 | $49.0 /1M chars | |
| 15 | 15-18 |  Soniox TTS Real-Time v2 | 1175±1295% interval 1163 to 1187 | 4,776 | Aug 2026 | $14.2 /1M chars | |
| 16 | 15-18 | Eleven v3 | 1171±1195% interval 1160 to 1182 | 4,481 | Feb 2026 | $100.0 /1M chars | |
| 17 | 15-18 |  Lightning V3.1 Pro (Jul 2026) | 1170±1495% interval 1156 to 1184 | 2,092 | Jul 2026 | $19.5 /1M chars | |
| 18 | 15-18 |  Speech 2.8 HD | 1170±1195% interval 1159 to 1181 | 4,560 | Feb 2026 | $100.0 /1M chars | |
| 19 | 19-21 |  Falcon 2 | 1154±1495% interval 1140 to 1168 | 1,868 | Jul 2026 | $10.0 /1M chars | |
| 20 | 19-21 | Speech 2.8 Turbo | 1152±1195% interval 1141 to 1163 | 4,426 | Feb 2026 | $60.0 /1M chars | |
| 21 | 19-22 |  Gradium TTS (Aug 2026) | 1149±1495% interval 1135 to 1163 | 1,801 | Aug 2026 | $47.2 /1M chars | |
| 22 | 21-27 |  Fish Audio S2.1 Pro | 1137±1395% interval 1124 to 1150 | 2,291 | Jun 2026 | $15.0 /1M chars | |
| 23 | 22-29 |  Async Flash v1.5 | 1133±1395% interval 1120 to 1146 | 2,202 | May 2026 | $10.1 /1M chars | |
| 24 | 22-30 |  SpaceXAI TTS | 1132±1595% interval 1117 to 1147 | 1,315 | Mar 2026 | $15.0 /1M chars | |
| 25 | 22-30 | Step TTS 2 (Mar 2026) | 1130±1395% interval 1117 to 1143 | 2,310 | Feb 2026 | $40.0 /1M chars | |
| 26 | 22-31 |  Azure HD 2.5 | 1129±1295% interval 1117 to 1141 | 2,508 | Mar 2026 | $22.0 /1M chars | |
| 27 | 22-30 | Speech 2.6 HD | 1128±1195% interval 1117 to 1139 | 3,911 | Oct 2025 | $100.0 /1M chars | |
| 28 | 23-32 | Simba 3.0 | 1123±1295% interval 1111 to 1135 | 2,568 | Feb 2026 | $6.6 /1M chars | |
| 29 | 25-32 | Speech 2.6 Turbo | 1121±1195% interval 1110 to 1132 | 4,227 | Oct 2025 | $60.0 /1M chars | |
| 30 | 25-32 | Async Pro v1.0 | 1118±1395% interval 1105 to 1131 | 2,102 | May 2026 | $20.2 /1M chars | |
| 31 | 26-32 | Fish Audio S2 Pro | 1117±1295% interval 1105 to 1129 | 2,325 | Mar 2026 | $15.0 /1M chars | |
| 32 | 29-34 | Speech-02-HD | 1111±1195% interval 1100 to 1122 | 4,321 | Mar 2025 | $100.0 /1M chars | |
| 33 | 32-38 |  TTS-1 HD | 1102±1295% interval 1090 to 1114 | 3,518 | Nov 2023 | $30.0 /1M chars | |
| 34 | 32-40 |  Chatterbox HD | 1101±1395% interval 1088 to 1114 | 2,363 | — | May 2025 | $40.0 /1M chars |
| 35 | 32-40 |  Speech 2.6 | 1100±1395% interval 1087 to 1113 | 2,057 | Jul 2026 | N/A | |
| 36 | 33-40 | Turbo v2.5 | 1096±1195% interval 1085 to 1107 | 6,786 | Jul 2024 | $50.0 /1M chars | |
| 37 | 33-42 | Step Audio EditX (Mar 2026) | 1092±1395% interval 1079 to 1105 | 2,012 | Feb 2026 | N/A | |
| 38 | 33-40 | Multilingual v2 | 1092±1095% interval 1082 to 1102 | 7,234 | Aug 2023 | $100.0 /1M chars | |
| 39 | 35-43 | TTS-1 | 1089±1195% interval 1078 to 1100 | 6,078 | Nov 2023 | $15.0 /1M chars | |
| 40 | 35-44 | Speech-02-Turbo | 1089±1295% interval 1077 to 1101 | 3,748 | Mar 2025 | $60.0 /1M chars | |
| 41 | 37-48 | Gemini 2.5 Flash Lite TTS | 1082±1295% interval 1070 to 1094 | 3,103 | Dec 2025 | $9.2 /1M chars | |
| 42 | 37-48 |  Voxtral TTS | 1079±1395% interval 1066 to 1092 | 2,094 | Mar 2026 | $16.0 /1M chars | |
| 43 | 39-48 | OpenAudio S1 | 1078±1195% interval 1067 to 1089 | 4,442 | Jun 2025 | $15.0 /1M chars | |
| 44 | 40-48 | Studio | 1077±1295% interval 1065 to 1089 | 4,389 | Feb 2023 | $160.0 /1M chars | |
| 45 | 41-48 | Flash v2.5 | 1076±1195% interval 1065 to 1087 | 5,643 | Dec 2024 | $50.0 /1M chars | |
| 46 | 41-52 | GPT-Realtime-2 | 1074±1395% interval 1061 to 1087 | 2,022 | May 2026 | $191.6 /1M chars | |
| 47 | 41-52 | Journey | 1073±1395% interval 1060 to 1086 | 2,186 | Dec 2023 | $160.0 /1M chars | |
| 48 | 41-54 |  Maya 2 Flash | 1071±1495% interval 1057 to 1085 | 1,867 | Jul 2026 | N/A | |
| 49 | 46-57 | T2A-01-HD | 1063±1195% interval 1052 to 1074 | 4,896 | Jan 2025 | $50.0 /1M chars | |
| 50 | 46-57 |  Kokoro 82M v1.0 | 1063±1195% interval 1052 to 1074 | 5,263 | Jan 2025 | $0.7 /1M chars | |
| 51 | 46-57 |  Polly Generative | 1063±1295% interval 1051 to 1075 | 3,681 | May 2024 | $30.0 /1M chars | |
| 52 | 46-57 |  Magpie-Multilingual 357M (Feb 2026) | 1062±1395% interval 1049 to 1075 | 2,037 | Feb 2026 | N/A | |
| 53 | 48-59 |  Coda | 1060±1395% interval 1047 to 1073 | 2,216 | May 2026 | $50.0 /1M chars | |
| 54 | 49-61 | Sonic 3 | 1058±1295% interval 1046 to 1070 | 3,429 | Oct 2025 | $49.0 /1M chars | |
| 55 | 49-62 |  MiMo-V2.5-TTS | 1054±1395% interval 1041 to 1067 | 1,918 | - | N/A | |
| 56 | 49-62 | Gemini 2.5 Flash TTS (Dec 2025) | 1053±1295% interval 1041 to 1065 | 2,815 | Dec 2025 | $9.2 /1M chars | |
| 57 | 49-62 | Chirp 3: HD | 1053±1295% interval 1041 to 1065 | 3,296 | Mar 2025 | $30.0 /1M chars | |
| 58 | 53-62 | Async Flash v1.0 | 1049±1295% interval 1037 to 1061 | 3,962 | Jul 2025 | $10.1 /1M chars | |
| 59 | 53-63 |  Octave 2 | 1048±1295% interval 1036 to 1060 | 3,053 | Oct 2025 | $87.5 /1M chars | |
| 60 | 53-65 | Maya 2 Global | 1047±1495% interval 1033 to 1061 | 1,749 | Jul 2026 | N/A | |
| 61 | 54-64 | Maya1 | 1046±1295% interval 1034 to 1058 | 3,218 | Jun 2025 | N/A | |
| 62 | 52-71 | OpenAudio S1 Mini | 1042±2095% interval 1022 to 1062 | 1,682 | Jun 2025 | $15.0 /1M chars | |
| 63 | 58-70 |  Higgs Audio V3 TTS | 1036±1395% interval 1023 to 1049 | 2,015 | Jun 2026 | $15.0 /1M chars | |
| 64 | 59-71 |  Bland Speech v3 | 1035±1395% interval 1022 to 1048 | 2,296 | Aug 2026 | $40.0 /1M chars | |
| 65 | 58-73 | Azure Neural | 1033±1995% interval 1014 to 1052 | 1,158 | Sept 2018 | $15.0 /1M chars | |
| 66 | 62-72 | Octave TTS | 1032±1295% interval 1020 to 1044 | 4,619 | Feb 2025 | $87.5 /1M chars | |
| 67 | 62-73 | MAI-Voice-1 | 1032±1395% interval 1019 to 1045 | 2,181 | Apr 2026 | N/A | |
| 68 | 62-73 | Polly Long-Form | 1031±1395% interval 1018 to 1044 | 1,893 | Nov 2023 | $100.0 /1M chars | |
| 69 | 62-73 | Gemini 2.5 Pro (Dec 2025) | 1030±1395% interval 1017 to 1043 | 2,409 | Dec 2025 | $18.3 /1M chars | |
| 70 | 63-73 | T2A-01-Turbo | 1028±1195% interval 1017 to 1039 | 4,918 | Jan 2025 | $30.0 /1M chars | |
| 71 | 65-73 | Chatterbox  | 1023±1295% interval 1011 to 1035 | 4,609 | May 2025 | $25.0 /1M chars | |
| 72 | 65-73 | Lightning v3.1 | 1021±1395% interval 1008 to 1034 | 2,145 | Mar 2026 | $25.0 /1M chars | |
| 73 | 65-73 | MiMo-V2-TTS | 1020±1495% interval 1006 to 1034 | 1,743 | Mar 2026 | N/A | |
| 74 | 74-75 |  Raon SpeechLM | 1002±1495% interval 988 to 1016 | 2,042 | Apr 2026 | N/A | |
| 75 | 75 |  Zonos-v0.1 | 1000AnchorFixed at 1000 as the reference point for this board's Elo scale. | 4,804 | Feb 2025 | $20.0 /1M chars | |
| 76 | 76 | Murf Speech Gen 2 | 982±1295% interval 970 to 994 | 4,965 | Jul 2024 | $100.0 /1M chars | |
| 77 | 77-79 | VibeVoice 1.5B | 955±1495% interval 941 to 969 | 2,095 | Aug 2025 | N/A | |
| 78 | 77-79 |  OpenVoice v2 | 952±1395% interval 939 to 965 | 2,970 | Apr 2024 | $8.3 /1M chars | |
| 79 | 77-82 | Qwen3 TTS Flash | 943±1495% interval 929 to 957 | 2,116 | Sept 2025 | $11.0 /1M chars | |
| 80 | 79-82 |  Neuphonic TTS | 937±1495% interval 923 to 951 | 2,373 | Oct 2024 | $20.8 /1M chars | |
| 81 | 79-82 |  Kugel 3 | 934±1395% interval 921 to 947 | 4,145 | Jun 2026 | N/A | |
| 82 | 80-82 | Qwen3 TTS | 930±1395% interval 917 to 943 | 2,893 | Jan 2026 | $13.8 /1M chars | |
| 83 | 83-85 | WaveNet | 915±1295% interval 903 to 927 | 6,229 | Sept 2016 | $4.0 /1M chars | |
| 84 | 82-85 |  XTTS v2 | 915±1595% interval 900 to 930 | 2,530 | — | Nov 2023 | $40.4 /1M chars |
| 85 | 83-85 | Mist V2 | 909±1595% interval 894 to 924 | 2,833 | Mar 2025 | $31.1 /1M chars | |
| 86 | 86-90 |  StyleTTS 2 | 893±1595% interval 878 to 908 | 2,440 | — | Jun 2023 | $2.8 /1M chars |
| 87 | 86-90 | Polly Neural | 893±1495% interval 879 to 907 | 2,741 | Jul 2019 | $16.0 /1M chars | |
| 88 | 86-90 | Neural2 | 892±1295% interval 880 to 904 | 6,452 | Jun 2022 | $16.0 /1M chars | |
| 89 | 86-90 | Standard | 883±1295% interval 871 to 895 | 6,500 | Aug 2018 | $4.0 /1M chars | |
| 90 | 86-90 |  Noiz TTS | 883±1595% interval 868 to 898 | 1,998 | Nov 2025 | N/A | |
| 91 | 91-92 |  MetaVoice v1 | 846±1895% interval 828 to 864 | 1,354 | — | Feb 2024 | N/A |
| 92 | 92 | Polly Standard | 828±1695% interval 812 to 844 | 3,295 | 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 (Elo 1320), 2. Sonic 3.6 (Elo 1273), 3. Gemini 3.8 Flash TTS (Elo 1272), 4. Qwen-Audio-3.0-TTS-Plus (Elo 1259), and 5. Realtime TTS-2 (Elo 1247). 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. Breeze TTS 2 (Elo 1206), 2. Fish Audio S2 Pro (Elo 1117), 3. Step Audio EditX (Mar 2026) (Elo 1092), 4. Voxtral TTS (Elo 1079), and 5. Kokoro 82M v1.0 (Elo 1063).