Project AURIONAI Use-Cases Research & Innovation
Artificial Analysis

Artificial Analysis Text-to-Speech Arena

Elo from blind preference votes on speech outputs; 96 models, 16 open weights

generative mediaactive

Score in Elo, AA Text-to-Speech arena, 96 models

  1. 1ElevenLabsEleven v4 Turbo1,327 Eloindependent
  2. 2ElevenLabsEleven v41,323 Eloindependent
  3. 3Alibaba (Qwen)Qwen-Audio-3.1-TTS-Plus1,298 Eloindependent
13scores
13models
13independent
0maintainer-run
0vendor-reported

Definition

Version
not in the notes
Category
generative media
Maintainer
Artificial Analysis
Status
active
Home page
https://artificialanalysis.ai/text-to-speech/leaderboard

Sources

  1. Text to Speech Leaderboard - Top AI Speech Models | Artificial Analysis artificialanalysis.ai https://artificialanalysis.ai/text-to-speech/leaderboard

Scores at a Glance

Score in Elo, AA Text-to-Speech arena, 96 models

Higher is better for this unit. Elo has no zero, so the axis starts at 1,000.

1ElevenLabsEleven v4 TurboElevenLabs
1,327 Elo
independent2ElevenLabsEleven v4ElevenLabs
1,323 Elo
independent3Alibaba (Qwen)Qwen-Audio-3.1-TTS-PlusAlibaba (Qwen)
1,298 Elo
independent4CartesiaSonic 3.6Cartesia
1,284 Elo
independent5GoogleGemini 3.8 Flash TTSGoogle
1,272 Elo
independent6Realtime TTS-2
1,257 Elo
independent7GoogleGemini 3.8 Flash-Lite TTSGoogle
1,244 Elo
independent8Simba 3.2
1,243 Elo
independent9BreezeBlueBreeze TTS 2BreezeBlue
1,222 Elo
independent10StepAudio 2.5 TTS
1,209 Elo
independent11ElevenLabsEleven v3ElevenLabs
1,176 Elo
independent12MiniMaxSpeech 2.8 HDMiniMax
1,173 Elo
independent13SpaceXAI (xAI)SpaceXAI TTSSpaceXAI (xAI)
1,143 Elo
independent
1,0001,1001,2001,3001,400
Where this comes from
Each bar is one model, with its most independent recorded score. Hatched bars are vendor-reported; the word beside each bar says who ran it. Settings and dates differ between rows; the Scores table below has every result.

Scores

13 of 13 scores shown aa-tts-scores.csv
ModelScoreWho ran itSettingView or populationDateSource
ElevenLabsEleven v4 Turbo1,327 Eloindependent–rank 1 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
ElevenLabsEleven v41,323 Eloindependent–rank 2 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
Alibaba (Qwen)Qwen-Audio-3.1-TTS-Plus1,298 Eloindependent–rank 3 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
CartesiaSonic 3.61,284 Eloindependent–rank 4 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
GoogleGemini 3.8 Flash TTS1,272 Eloindependent–rank 5 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
Realtime TTS-21,257 Eloindependent–rank 7 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
GoogleGemini 3.8 Flash-Lite TTS1,244 Eloindependent–rank 8 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
Simba 3.21,243 Eloindependent–rank 9 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
BreezeBlueBreeze TTS 21,222 Eloindependent–rank 10 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
StepAudio 2.5 TTS1,209 Eloindependent–rank 14 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
ElevenLabsEleven v31,176 Eloindependent–rank 18 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
MiniMaxSpeech 2.8 HD1,173 Eloindependent–rank 19 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
SpaceXAI (xAI)SpaceXAI TTS1,143 Eloindependent–rank 25 of 96, AA Text-to-Speech arena, 96 models10 Oct 2026artificialanalysis.ai
Where this comes from
Scores from different settings, harnesses or dates are not directly comparable; the row says which. Sorted by score, highest first, as a reading aid, not a ranking.

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "Artificial Analysis Text-to-Speech Arena", https://aurion.semai.id/benchmarks/aa-tts/, records of October 2026.