Bestenliste der Arena für kontrollierte StimmenArtificial Analysis
Vergleichen Sie TTS-Modelle mit denselben 8 geklonten Stimmen (4 US, 4 UK).
Range | Creator | Model | Elo | 95% CI | Samples | Veröffentlicht | API-Preise 1 | |
|---|---|---|---|---|---|---|---|---|
| 1 | 1 | Sonic 3.6 | 1,119 | -15/15 | 1,993 | Aug 2026 | 49,0 $ /1 Mio. Zeichen | |
| 2 | 2 | Sonic 3.5 | 1,096 | -12/12 | 3,791 | May 2026 | 49,0 $ /1 Mio. Zeichen | |
| 3 | 3-4 | Realtime TTS-2 Flash - Research Preview | 1,075 | -15/15 | 1,621 | Aug 2026 | 10,4 $ /1 Mio. Zeichen | |
| 4 | 4-5 | Eleven v3 | 1,062 | -12/12 | 3,244 | Feb 2026 | 100,0 $ /1 Mio. Zeichen | |
| 5 | 4-7 | StepAudio 2.5 TTS | 1,052 | -13/13 | 2,558 | Apr 2026 | 85,0 $ /1 Mio. Zeichen | |
| 6 | 5-8 | Realtime TTS 1.5 Max | 1,047 | -14/14 | 2,488 | Jan 2026 | 26,0 $ /1 Mio. Zeichen | |
| 7 | 5-8 | Realtime TTS-2 - Research Preview | 1,045 | -13/13 | 2,800 | May 2026 | 20,8 $ /1 Mio. Zeichen | |
| 8 | 6-9 | SpaceXAI TTS | 1,039 | -13/13 | 2,524 | Mar 2026 | 15,0 $ /1 Mio. Zeichen | |
| 9 | 8-10 | Speech 2.8 HD | 1,028 | -12/12 | 3,188 | Feb 2026 | 100,0 $ /1 Mio. Zeichen | |
| 10 | 8-12 | v3 Conversational | 1,026 | -14/14 | 1,626 | Aug 2026 | 50,0 $ /1 Mio. Zeichen | |
| 11 | 10-16 | Speech 2.8 Turbo | 1,015 | -12/12 | 3,207 | Feb 2026 | 60,0 $ /1 Mio. Zeichen | |
| 12 | 10-17 | Simba 3.0 | 1,015 | -13/13 | 2,553 | Feb 2026 | 12,5 $ /1 Mio. Zeichen | |
| 13 | 11-22 | Soniox TTS Real-Time v1 | 1,008 | -12/12 | 2,935 | Jul 2026 | 14,2 $ /1 Mio. Zeichen | |
| 14 | 11-23 | Breeze TTS 2 | 1,007 | -15/15 | 1,546 | Aug 2026 | 34,0 $ /1 Mio. Zeichen | |
| 15 | 11-23 | Voxtral TTS | 1,005 | -13/13 | 2,554 | Mar 2026 | 16,0 $ /1 Mio. Zeichen | |
| 16 | 11-23 | Multilingual v2 | 1,005 | -12/12 | 3,140 | Aug 2023 | 100,0 $ /1 Mio. Zeichen | |
| 17 | 12-23 | Fish Audio S2 Pro | 1,003 | -12/12 | 3,765 | Mar 2026 | 15,0 $ /1 Mio. Zeichen | |
| 18 | 18 | OpenAudio S1 | 1,000 | 0/0 | 5,013 | Jun 2025 | 15,0 $ /1 Mio. Zeichen | |
| 19 | 13-23 | OpenAudio S1 Mini | 1,000 | -13/13 | 2,592 | Jun 2025 | 15,0 $ /1 Mio. Zeichen | |
| 20 | 13-23 | Gradium TTS Beta | 1,000 | -14/14 | 2,134 | Jul 2026 | 47,2 $ /1 Mio. Zeichen | |
| 21 | 13-23 | Fish Audio S2.1 Pro | 999 | -13/13 | 2,556 | Jun 2026 | 15,0 $ /1 Mio. Zeichen | |
| 22 | 13-23 | Realtime TTS 1.5 Mini | 999 | -13/13 | 2,542 | Jan 2026 | 10,4 $ /1 Mio. Zeichen | |
| 23 | 16-23 | Fish Audio S1 | 993 | -12/12 | 3,398 | Oct 2025 | 15,0 $ /1 Mio. Zeichen | |
| 24 | 24-28 | Flash v2.5 | 974 | -12/12 | 3,204 | Dec 2024 | 50,0 $ /1 Mio. Zeichen | |
| 25 | 24-29 | Turbo v2.5 | 970 | -13/13 | 2,796 | Jul 2024 | 50,0 $ /1 Mio. Zeichen | |
| 26 | 24-29 | Gradium TTS | 970 | -13/13 | 2,621 | Jun 2026 | 47,2 $ /1 Mio. Zeichen | |
| 27 | 24-29 | Kugel 3 | 968 | -12/12 | 4,728 | - | k. A. | |
| 28 | 24-29 | Lightning V3.1 Pro TTS (Jun 2026) | 965 | -13/13 | 2,745 | Jun 2026 | 19,5 $ /1 Mio. Zeichen | |
| 29 | 25-29 | Higgs Audio V3 TTS | 957 | -13/13 | 2,448 | Jun 2026 | 15,0 $ /1 Mio. Zeichen | |
| 30 | 30-33 | Sonic 3 | 936 | -13/13 | 2,988 | Oct 2025 | 49,0 $ /1 Mio. Zeichen | |
| 31 | 30-35 | Chatterbox | 931 | -13/13 | 2,865 | May 2025 | 25,0 $ /1 Mio. Zeichen | |
| 32 | 30-36 | Magpie TTS ZeroshotOpen Weights | 928 | -16/16 | 1,061 | Jul 2026 | k. A. | |
| 33 | 30-36 | Noiz TTS | 924 | -13/13 | 2,296 | Nov 2025 | k. A. | |
| 34 | 31-37 | VibeVoice 7B | 919 | -13/13 | 2,809 | Aug 2025 | k. A. | |
| 35 | 31-37 | Raon SpeechLM | 918 | -15/15 | 1,619 | - | k. A. | |
| 36 | 31-37 | Soniox TTS Real-Time v2 | 917 | -15/15 | 1,590 | Aug 2026 | 14,2 $ /1 Mio. Zeichen | |
| 37 | 34-37 | LMNT | 907 | -13/13 | 2,637 | Dec 2023 | 49,0 $ /1 Mio. Zeichen | |
| 38 | 38 | Lightning v3.1 | 886 | -14/14 | 2,500 | Mar 2026 | 25,0 $ /1 Mio. Zeichen | |
| 39 | 39 | XTTS v2 | 824 | -14/14 | 2,807 | Nov 2023 | 40,4 $ /1 Mio. Zeichen | |
| 40 | 40 | OpenVoice v2 | 792 | -16/16 | 2,487 | Apr 2024 | 8,3 $ /1 Mio. Zeichen |
1 API-Preise spiegeln die Kosten wider, 1M Zeichen über die API des Modellerstellers mit den Standardeinstellungen des Modells zu erzeugen