Skip to content
IFAI

Model rankings

1037 models · 5 boards · 2 sources

‹
Week of October 19, 2026
November 9, 202636 changesNovember 2, 202641 changesOctober 26, 20269 changesOctober 19, 2026247 changesOctober 12, 2026127 changesOctober 5, 202634 changesSeptember 28, 202617 changesSeptember 21, 2026115 changesSeptember 14, 2026174 changesSeptember 7, 2026144 changesAugust 31, 202620 changesAugust 24, 202615 changes
›

Text-to-speech arena · Anonymous head-to-head votes · Updated October 19, 2026

Best voice
Eleven v4 TurboElevenLabs1,332Score
Best open model
Fish Audio Drama 3 PreviewFish Audio1,247Score
Best from China
Qwen-Audio-3.1-TTS-PlusAlibaba1,297Score
Top three by Score
1Eleven v4 TurboElevenLabs1,332
2Eleven v4ElevenLabs1,329
3Qwen-Audio-3.1-TTS-PlusAlibaba1,297
Top three open models
1Fish Audio Drama 3 PreviewFish Audio · #9 overall1,247
2Breeze TTS 2BreezeBlue · #11 overall1,222
3Ming-3.0-ttsAnt Group · #18 overall1,188
This week's changes
Muse Wonder Full (Max)Intelligence · #3New
FLUX 3 Max (720p)Video · #4New
Muse Wonder Full (Max)Coding · #6New
Muse Wonder (Max)Intelligence · #7New
Kimi K3.1 Pro (Max)Coding · #7New
AllOpenClosed
All regionsUSChina
One row per modelAll variants
RankChangeModelScore95% interval
1—Eleven v4 TurboClosedElevenLabs1,332±17
2—Eleven v4ClosedElevenLabs1,329±16
3—Qwen-Audio-3.1-TTS-PlusClosedAlibaba1,297±16
4—Sonic 3.6ClosedCartesia1,291±14
5—Gemini 3.8 Flash TTSClosedGoogle1,280±14
6—Qwen-Audio-3.0-TTS-PlusClosedAlibaba1,264±14
7—Realtime TTS-2ClosedInworld1,262±15
8—Gemini 3.8 Flash-Lite TTSClosedGoogle1,252±14
9NewFish Audio Drama 3 PreviewOpenFish Audio1,247±15
10▼ 1Simba 3.2ClosedSpeechifyAI1,245±12
11▼ 1Breeze TTS 2OpenBreezeBlue1,222±14
12▲ 2StepAudio 2.5 TTS (Aug 2026)ClosedStepFun1,217±14
13▼ 2Realtime TTS-2 FlashClosedInworld1,217±13
14▼ 2Luna TTSClosedVUI Labs1,213±12
15▼ 2Gemini 3.1 Flash TTSClosedGoogle1,212±11
16▼ 1v3 ConversationalClosedElevenLabs1,212±13
17▼ 1Sonic 3.5ClosedCartesia1,191±11
18NewMing-3.0-ttsOpenAnt Group1,188±14
19NewFish Audio S2.1 Pro (Oct 2026)ClosedFish Audio1,187±14
20▼ 2Eleven v3ClosedElevenLabs1,182±11
21▼ 4Soniox TTS Real-Time v2ClosedSoniox1,179±11
22▼ 2Gradium TTS (Sep 2026)ClosedGradium1,175±13
23▼ 4Speech 2.8 HDClosedMiniMax1,174±11
24▼ 3Lightning V3.1 Pro (Jul 2026)ClosedSmallest.ai1,168±12
25▼ 1Speech 2.8 TurboClosedMiniMax1,155±11
26▼ 4Falcon 2ClosedMurf AI1,153±13
27▼ 4Gradium TTS (Aug 2026)ClosedGradium1,152±13
28▼ 3SpaceXAI TTSClosedSpaceXAI1,149±13
29—Azure HD 2.5ClosedMicrosoft1,140±12
30▼ 2Async Flash v1.5ClosedAsync1,139±12
31▼ 4Step TTS 2 (Mar 2026)ClosedStepFun1,135±12
32▼ 2Speech 2.6 HDClosedMiniMax1,129±11
33▼ 2Async Pro v1.0ClosedAsync1,127±12
34▼ 2Simba 3.0ClosedSpeechifyAI1,121±11
35▼ 2Speech 2.6 TurboClosedMiniMax1,121±11
36▼ 2Fish Audio S2 ProOpenFish Audio1,113±12
37▼ 2Speech-02-HDClosedMiniMax1,111±11
38▼ 1Chatterbox HDClosedResemble AI1,109±12
39▼ 3TTS-1 HDClosedOpenAI1,102±11
40—Step Audio EditX (Mar 2026)OpenStepFun1,099±12
41▼ 2Turbo v2.5ClosedElevenLabs1,098±10
42▼ 4Speech 2.6ClosedHithink1,096±12
43▼ 2Multilingual v2ClosedElevenLabs1,096±10
44▼ 2Speech-02-TurboClosedMiniMax1,091±11
45▼ 1Voxtral TTSOpenMistral1,091±12
46▼ 3TTS-1ClosedOpenAI1,089±10
47▼ 2Gemini 2.5 Flash Lite TTSClosedGoogle1,086±11
48▼ 2OpenAudio S1ClosedFish Audio1,085±11
49▼ 2StudioClosedGoogle1,077±11
50▼ 1Flash v2.5ClosedElevenLabs1,077±11

Intelligence over time

Each maker's best model at the time on the Artificial Analysis Intelligence Index, by the month it was released.

Gained most · Anthropic +57.2Highest now · Anthropic 64.6Gap between highest and lowest 18.1 ptsTime range Nov 2025 – Nov 2026
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi

Arena data comes from the leaderboard dataset it publishes on Hugging Face (CC BY 4.0). Only each board's overall leaderboard is used; scores and ranks are unchanged. dataset · CC BY 4.0

Artificial Analysis data comes from its API (artificialanalysis.ai). Scores are unchanged; ranks follow the scores. artificialanalysis.ai

Open or closed follows the license Arena gives; prices and Artificial Analysis intelligence show only where model names match, otherwise “—”.