Ranking da Arena de voz do provedorArtificial Analysis
Compare modelos de texto para fala (TTS) usando as vozes nativas de cada provedor.
Range | Creator | Model | Elo | 95% CI | Samples | Vozes da Arena | Lançamento | Preço da API 1 | |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 1 | Breeze TTS 2 | 1205 | -17/17 | 1,227 | Aug 2026 | US$ 34,0 /1 mi de caracteres | ||
| 2 | 2 | Fish Audio S2 Pro | 1122 | -13/13 | 2,257 | Mar 2026 | US$ 15,0 /1 mi de caracteres | ||
| 3 | 3 | Step Audio EditX (Mar 2026) | 1098 | -13/13 | 2,004 | Feb 2026 | N/D | ||
| 4 | 4 | Voxtral TTS | 1074 | -13/13 | 2,104 | Mar 2026 | US$ 16,0 /1 mi de caracteres | ||
| 5 | 5-6 | Kokoro 82M v1.0 | 1059 | -11/11 | 5,278 | Jan 2025 | US$ 0,7 /1 mi de caracteres | ||
| 6 | 5-6 | Magpie-Multilingual 357M (Feb 2026) | 1057 | -13/13 | 2,000 | Feb 2026 | N/D | ||
| 7 | 5-9 | OpenAudio S1 Mini | 1040 | -20/20 | 1,681 | Jun 2025 | US$ 15,0 /1 mi de caracteres | ||
| 8 | 7-9 | Maya1 | 1037 | -12/12 | 3,244 | Jun 2025 | N/D | ||
| 9 | 7-9 | Higgs Audio V3 TTS | 1034 | -13/13 | 2,000 | Jun 2026 | US$ 15,0 /1 mi de caracteres | ||
| 10 | 10 | Chatterbox | 1019 | -11/11 | 4,638 | May 2025 | US$ 25,0 /1 mi de caracteres | ||
| 11 | 11 | Zonos-v0.1 | 1000 | 0/0 | 4,860 | Feb 2025 | US$ 20,0 /1 mi de caracteres | ||
| 12 | 12-13 | VibeVoice 1.5B | 960 | -14/14 | 2,129 | Aug 2025 | N/D | ||
| 13 | 12-13 | OpenVoice v2 | 951 | -13/13 | 2,978 | — | Apr 2024 | US$ 8,3 /1 mi de caracteres | |
| 14 | 14 | XTTS v2 | 914 | -15/15 | 2,543 | — | Nov 2023 | US$ 40,4 /1 mi de caracteres | |
| 15 | 15 | StyleTTS 2 | 893 | -15/15 | 2,449 | — | Jun 2023 | US$ 2,8 /1 mi de caracteres | |
| 16 | 16 | MetaVoice v1 | 835 | -18/18 | 1,368 | — | Feb 2024 | N/D |
1 O preço da API reflete o custo de gerar 1M de caracteres na API do criador do modelo com as configurações padrão do modelo