Ranking da Arena de voz controladaArtificial Analysis

Compare modelos de texto para fala (TTS) usando as mesmas 8 vozes clonadas (4 dos EUA, 4 do Reino Unido).

Range
Creator
Model
Elo
95% CI
Samples
Lançamento
Preço da API 1
11-2
Inworld logoInworld
Realtime TTS-2
1123-16/161,292Aug 2026US$ 20,8 /1 mi de caracteres
21-2
Cartesia logoCartesia
Sonic 3.6
1119-15/151,993Aug 2026US$ 49,0 /1 mi de caracteres
33
Cartesia logoCartesia
Sonic 3.5
1096-12/123,791May 2026US$ 49,0 /1 mi de caracteres
44-5
Inworld logoInworld
Realtime TTS-2 Flash - Research Preview
1075-15/151,621Aug 2026US$ 10,4 /1 mi de caracteres
55-6
ElevenLabs logoElevenLabs
Eleven v3
1062-12/123,244Feb 2026US$ 100,0 /1 mi de caracteres
65-6
StepFun logoStepFun
StepAudio 2.5 TTS
1052-13/132,558Apr 2026US$ 85,0 /1 mi de caracteres
77-8
SpaceXAI logoSpaceXAI
SpaceXAI TTS
1039-13/132,524Mar 2026US$ 15,0 /1 mi de caracteres
87-9
MiniMax logoMiniMax
Speech 2.8 HD
1028-12/123,188Feb 2026US$ 100,0 /1 mi de caracteres
97-11
ElevenLabs logoElevenLabs
v3 Conversational
1026-14/141,626Aug 2026US$ 50,0 /1 mi de caracteres
109-15
MiniMax logoMiniMax
Speech 2.8 Turbo
1015-12/123,207Feb 2026US$ 60,0 /1 mi de caracteres
119-16
SpeechifyAI logoSpeechifyAI
Simba 3.0
1015-13/132,553Feb 2026US$ 12,5 /1 mi de caracteres
1210-20
Soniox logoSoniox
Soniox TTS Real-Time v1
1008-12/122,935Jul 2026US$ 14,2 /1 mi de caracteres
1310-21
BreezeBlue logoBreezeBlue
Breeze TTS 2Hugging FaceOpen Weights
1007-15/151,546Aug 2026US$ 34,0 /1 mi de caracteres
1410-21
Mistral logoMistral
Voxtral TTSHugging FaceOpen Weights
1005-13/132,554Mar 2026US$ 16,0 /1 mi de caracteres
1510-21
ElevenLabs logoElevenLabs
Multilingual v2
1005-12/123,140Aug 2023US$ 100,0 /1 mi de caracteres
1611-21
Fish Audio logoFish Audio
Fish Audio S2 ProHugging FaceOpen Weights
1003-12/123,765Mar 2026US$ 15,0 /1 mi de caracteres
1717
Fish Audio logoFish Audio
OpenAudio S1
10000/05,013Jun 2025US$ 15,0 /1 mi de caracteres
1812-21
Fish Audio logoFish Audio
OpenAudio S1 Mini
1000-13/132,592Jun 2025US$ 15,0 /1 mi de caracteres
1912-21
Gradium logoGradium
Gradium TTS (Aug 2026)
1000-14/142,134Aug 2026US$ 47,2 /1 mi de caracteres
2012-21
Fish Audio logoFish Audio
Fish Audio S2.1 Pro
999-13/132,556Jun 2026US$ 15,0 /1 mi de caracteres
2115-21
Fish Audio logoFish Audio
Fish Audio S1
993-12/123,398Oct 2025US$ 15,0 /1 mi de caracteres
2222-25
ElevenLabs logoElevenLabs
Flash v2.5
974-12/123,204Dec 2024US$ 50,0 /1 mi de caracteres
2322-26
ElevenLabs logoElevenLabs
Turbo v2.5
970-13/132,796Jul 2024US$ 50,0 /1 mi de caracteres
2422-26
KugelAudio logoKugelAudio
Kugel 3
968-12/124,728Jun 2026N/D
2522-26
Smallest.ai logoSmallest.ai
Lightning V3.1 Pro TTS (Jun 2026)
965-13/132,745Jun 2026US$ 19,5 /1 mi de caracteres
2623-26
Boson AI logoBoson AI
Higgs Audio V3 TTSHugging FaceOpen Weights
957-13/132,448Jun 2026US$ 15,0 /1 mi de caracteres
2727-30
Cartesia logoCartesia
Sonic 3
936-13/132,988Oct 2025US$ 49,0 /1 mi de caracteres
2827-32
Resemble AI logoResemble AI
ChatterboxGitHubOpen Weights
931-13/132,865May 2025US$ 25,0 /1 mi de caracteres
2927-33
NVIDIA logoNVIDIA
Magpie TTS ZeroshotOpen Weights
928-16/161,061Jul 2026N/D
3027-33
Noiz logoNoiz
Noiz TTS
924-13/132,296Nov 2025N/D
3128-34
Microsoft logoMicrosoft
VibeVoice 7BHugging FaceOpen Weights
919-13/132,809Aug 2025N/D
3228-34
Krafton logoKrafton
Raon SpeechLM
918-15/151,619Apr 2026N/D
3328-34
Soniox logoSoniox
Soniox TTS Real-Time v2
917-15/151,590Aug 2026US$ 14,2 /1 mi de caracteres
3431-34
LMNT logoLMNT
LMNT
907-13/132,637Dec 2023US$ 49,0 /1 mi de caracteres
3535
Smallest.ai logoSmallest.ai
Lightning v3.1
886-14/142,500Mar 2026US$ 25,0 /1 mi de caracteres
3636
Coqui logoCoqui
824-14/142,807Nov 2023US$ 40,4 /1 mi de caracteres
3737
OpenVoice logoOpenVoice
OpenVoice v2Hugging FaceOpen Weights
792-16/162,487Apr 2024US$ 8,3 /1 mi de caracteres

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