Controlled Voice Arena LeaderboardArtificial Analysis
Compare Text to Speech (TTS) models using the same 8 cloned voices (4 US, 4 UK).
Range | Model | Elo | Samples | Released | API Pricing 1 | |
|---|---|---|---|---|---|---|
| 1 | 1-2 |  HeyGen Voice | 1199±1595% interval 1184 to 1214 | 1,938 | Oct 2026 | $30.0 /1M chars |
| 2 | 2 |  Qwen-Audio-3.1-TTS-Plus | 1185±1495% interval 1171 to 1199 | 2,272 | Sept 2026 | $19.3 /1M chars |
| 3 | 3-6 |  Eleven v4 | 1158±1395% interval 1145 to 1171 | 2,511 | Sept 2026 | $80.0 /1M chars |
| 4 | 3-6 | Eleven v4 Turbo | 1158±1395% interval 1145 to 1171 | 2,570 | Sept 2026 | $40.0 /1M chars |
| 5 | 3-6 |  Sonic 3.6 | 1151±1395% interval 1138 to 1164 | 3,162 | Aug 2026 | $49.0 /1M chars |
| 6 | 3-6 |  Realtime TTS-2 | 1147±1395% interval 1134 to 1160 | 3,116 | Aug 2026 | $20.8 /1M chars |
| 7 | 7-8 | Qwen-Audio-3.0-TTS-Plus | 1126±1395% interval 1113 to 1139 | 2,904 | Sept 2026 | $19.3 /1M chars |
| 8 | 7-8 | Sonic 3.5 | 1118±1295% interval 1106 to 1130 | 5,188 | May 2026 | $49.0 /1M chars |
| 9 | 9-10 |  Simba 3.2 | 1087±1395% interval 1074 to 1100 | 2,894 | Jul 2026 | $6.6 /1M chars |
| 10 | 9-12 |  StepAudio 2.5 TTS | 1078±1295% interval 1066 to 1090 | 3,827 | Apr 2026 | $85.0 /1M chars |
| 11 | 10-13 | Eleven v3 | 1073±1295% interval 1061 to 1085 | 4,427 | Feb 2026 | $100.0 /1M chars |
| 12 | 10-15 | Realtime TTS-2 Flash | 1069±1295% interval 1057 to 1081 | 3,239 | Aug 2026 | $10.4 /1M chars |
| 13 | 11-17 | v3 Conversational | 1063±1395% interval 1050 to 1076 | 2,804 | Aug 2026 | $50.0 /1M chars |
| 14 | 11-18 |  Fish Audio Drama 3 Preview  | 1059±1495% interval 1045 to 1073 | 2,150 | Oct 2026 | $15.0 /1M chars |
| 15 | 12-18 |  SpaceXAI TTS | 1057±1295% interval 1045 to 1069 | 3,446 | Mar 2026 | $15.0 /1M chars |
| 16 | 12-18 |  Gradium TTS (Sep 2026) | 1057±1395% interval 1044 to 1070 | 2,326 | Sept 2026 | $47.8 /1M chars |
| 17 | 12-18 |  Gemini 3.8 Flash-Lite TTS | 1057±1395% interval 1044 to 1070 | 3,119 | Sept 2026 | $11.0 /1M chars |
| 18 | 14-18 | Gemini 3.8 Flash TTS | 1049±1395% interval 1036 to 1062 | 3,036 | Sept 2026 | $16.5 /1M chars |
| 19 | 19-20 | Simba 3.0 | 1036±1395% interval 1023 to 1049 | 3,252 | Feb 2026 | $6.6 /1M chars |
| 20 | 20-23 |  Speech 2.8 HD | 1023±1295% interval 1011 to 1035 | 3,802 | Feb 2026 | $100.0 /1M chars |
| 21 | 20-24 | Speech 2.8 Turbo | 1022±1295% interval 1010 to 1034 | 4,040 | Feb 2026 | $60.0 /1M chars |
| 22 | 20-24 |  Breeze TTS 2 | 1018±1395% interval 1005 to 1031 | 2,659 | Aug 2026 | $34.0 /1M chars |
| 23 | 20-27 | Gradium TTS (Aug 2026) | 1015±1395% interval 1002 to 1028 | 2,669 | Aug 2026 | $47.2 /1M chars |
| 24 | 20-29 | Fish Audio S2.1 Pro (Oct 2026) | 1011±1495% interval 997 to 1025 | 2,011 | Oct 2026 | $15.0 /1M chars |
| 25 | 23-29 |  Voxtral TTS | 1005±1395% interval 992 to 1018 | 2,993 | Mar 2026 | $16.0 /1M chars |
| 26 | 23-29 | OpenAudio S1 Mini | 1004±1395% interval 991 to 1017 | 2,949 | Jun 2025 | $15.0 /1M chars |
| 27 | 24-29 | Fish Audio S2 Pro | 1002±1295% interval 990 to 1014 | 4,490 | Mar 2026 | $15.0 /1M chars |
| 28 | 28 | OpenAudio S1 | 1000AnchorFixed at 1000 as the reference point for this board's Elo scale. | 6,207 | Jun 2025 | $15.0 /1M chars |
| 29 | 25-30 | Multilingual v2 | 997±1295% interval 985 to 1009 | 3,791 | Aug 2023 | $100.0 /1M chars |
| 30 | 29-31 | Flash v2.5 | 986±1395% interval 973 to 999 | 3,748 | Dec 2024 | $50.0 /1M chars |
| 31 | 29-31 | Fish Audio S1 | 985±1395% interval 972 to 998 | 3,737 | Oct 2025 | $15.0 /1M chars |
| 32 | 32-33 | Turbo v2.5 | 971±1395% interval 958 to 984 | 2,880 | Jul 2024 | $50.0 /1M chars |
| 33 | 32-34 |  Kugel 3 | 963±1395% interval 950 to 976 | 4,152 | Jun 2026 | N/A |
| 34 | 32-35 |  Higgs Audio V3 TTS | 958±1495% interval 944 to 972 | 2,578 | Jun 2026 | $15.0 /1M chars |
| 35 | 33-37 | Sonic 3 | 949±1495% interval 935 to 963 | 2,825 | Oct 2025 | $49.0 /1M chars |
| 36 | 35-38 |  Soniox TTS Real-Time v2 | 943±1495% interval 929 to 957 | 2,351 | Aug 2026 | $14.2 /1M chars |
| 37 | 35-39 | 940±1795% interval 923 to 957 | 1,060 | Jul 2026 | N/A | |
| 38 | 36-40 |  Chatterbox  | 932±1495% interval 918 to 946 | 2,795 | May 2025 | $25.0 /1M chars |
| 39 | 37-40 |  Raon SpeechLM | 928±1495% interval 914 to 942 | 2,325 | Apr 2026 | N/A |
| 40 | 38-40 |  Noiz TTS | 919±1495% interval 905 to 933 | 2,364 | Nov 2025 | N/A |
| 41 | 41 |  Lightning v3.1 | 886±1595% interval 871 to 901 | 2,418 | Mar 2026 | $25.0 /1M chars |
| 42 | 42 |  XTTS v2 | 828±1695% interval 812 to 844 | 2,552 | Nov 2023 | $40.4 /1M chars |
| 43 | 43 |  OpenVoice v2 | 809±1795% interval 792 to 826 | 2,227 | Apr 2024 | $8.3 /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.