Speech Generation - Rankings
Compare speech generation models based on Artificial Analysis Arena.
Select a metric
ELO arena ranking for speech synthesis quality based on blind pairwise human preference comparisons.
Data from Artificial Analysis
Data from artificialanalysis.ai| # | ||
|---|---|---|
| 1 |  Inworld | 1,162±15 |
| 2 | Inworld | 1,115±19 |
| 3 |  OpenAI | 1,111±8 |
| 4 |  MiniMax | 1,107±11 |
| 5 | MiniMax | 1,105±12 |
| 6 | MiniMax | 1,105±11 |
| 7 |  ElevenLabs | 1,105±7 |
| 8 | MiniMax | 1,103±12 |
| 9 | ElevenLabs | 1,096±7 |
| 10 | Inworld | 1,095±10 |
| 11 | ElevenLabs | 1,095±11 |
| 12 | ElevenLabs | 1,088±8 |
| 13 |  StepFun | 1,075±20 |
| 14 |  Fish Audio | 1,070±9 |
| 15 | Fish Audio | 1,061±12 |
| 16 |  Kokoro | 1,060±9 |
| 17 | OpenAI | 1,059±15 |
| 18 |  Amazon | 1,057±10 |
| 19 | MiniMax | 1,056±9 |
| 20 |  Cartesia | 1,054±14 |
| 21 | Amazon | 1,053±19 |
| 22 |  Hume AI | 1,051±14 |
| 23 |  Microsoft Azure | 1,050±7 |
| 24 | Resemble AI | 1,048±9 |
| 25 |  Google | 1,045±10 |
| 26 | OpenAI | 1,044±16 |
| 27 |  async | 1,041±11 |
| 28 | Cartesia | 1,033±13 |
| 29 | MiniMax | 1,032±9 |
| 30 |  Speechify | 1,032±9 |
| 31 |  Maya Research | 1,030±14 |
| 32 | Google | 1,027±14 |
| 33 | Hume AI | 1,027±9 |
| 34 | 1,026±18 | |
| 35 | Fish Audio | 1,026±8 |
| 36 | Google | 1,025±18 |
| 37 | Google | 1,023±11 |
| 38 | Google | 1,022±11 |
| 39 | StepFun | 1,020±47 |
| 40 | 1,013±19 | |
| 41 | Resemble AI | 1,013±9 |
| 42 |  NVIDIA | 1,006±12 |
| 43 | 1,004±18 | |
| 44 |  Zyphra | 1,000±0 |
| 45 | NVIDIA | 992±16 |
| 46 |  Murf AI | 987±8 |
| 47 |  LMNT | 982±10 |
| 48 | Microsoft Azure | 980±17 |
| 49 |  Alibaba | 974±18 |
| 50 |  OpenVoice | 973±7 |
| 51 | Microsoft Azure | 970±17 |
| 52 | StepFun | 959±10 |
| 53 |  Neuphonic | 950±20 |
| 54 | Murf AI | 943±27 |
| 55 | Amazon | 942±14 |
| 56 |  Coqui | 910±8 |
| 57 | Google | 903±7 |
| 58 |  StyleTTS | 901±8 |
| 59 | Google | 874±7 |
| 60 | Google | 871±7 |
| 61 | Amazon | 843±11 |
| 62 |  MetaVoice | 825±13 |
Last updated Feb 26, 2026