TTS
| # | Voice | Overall | Naturalness · MOS | Time to first audio · faster ▸ | Cost / 1M |
|---|---|---|---|---|---|
| 1 | sonic-3.5Cartesia | 94 | 3.42 | 119ms | $39.0 |
| 2 | aura-2Deepgram | 94 | 3.47 | 195ms | $30.0 |
| 3 | inworld-tts-2Inworld | 92 | 3.34 | 137ms | $20.8 |
| 4 | eleven_v3ElevenLabs | 83 | 3.51 | 487ms | $100 |
| 5 | tts-rt-v1Soniox | 83 | 3.39 | 398ms | — |
| 6 | lightning_v3.1Smallest | 82 | 3.25 | 316ms | $25.0 |
| 7 | grok-ttsxAI Grok | 71 | 2.96 | 344ms | $15.0 |
| 8 | simba-3.2Speechify | 70 | 3.41 | 704ms | $10.0 |
| 9 | speech-2.8-hdMiniMax | 61 | 2.60 | 309ms | $100.0 |
| 10 | arcanav3Rime | 60 | 2.59 | 318ms | $40.0 |
| 11 | defaultGradium | 57 | 2.69 | 465ms | $47.2 |
| 12 | gemini-3.1-flash-tts-previewGemini | 57 | 3.51 | 1093ms | $75.0 |
| 13 | gpt-4o-mini-ttsOpenAI | 44 | 2.56 | 674ms | — |
| 14 | octave-2Hume | 38 | 2.30 | 609ms | $87.5 |
| 15 | qwen3-tts-flashQwen | 33 | 1.98 | 491ms | $10.0 |
Naturalnesshigher is better
eleven_v33.51
gemini-3.1-flash-tts-preview3.51
aura-23.47
sonic-3.53.42
simba-3.23.41
tts-rt-v13.39
inworld-tts-23.34
lightning_v3.13.25
grok-tts2.96
default2.69
speech-2.8-hd2.60
arcanav32.59
gpt-4o-mini-tts2.56
octave-22.30
qwen3-tts-flash1.98
Time to first audiofaster is better
sonic-3.5119ms
inworld-tts-2137ms
aura-2195ms
speech-2.8-hd309ms
lightning_v3.1316ms
arcanav3318ms
grok-tts344ms
tts-rt-v1398ms
default465ms
eleven_v3487ms
qwen3-tts-flash491ms
octave-2609ms
gpt-4o-mini-tts674ms
simba-3.2704ms
gemini-3.1-flash-tts-preview1093ms
Cost / 1M charscheaper is better
simba-3.2$10.0
qwen3-tts-flash$10.0
grok-tts$15.0
inworld-tts-2$20.8
lightning_v3.1$25.0
aura-2$30.0
sonic-3.5$39.0
arcanav3$40.0
default$47.2
gemini-3.1-flash-tts-preview$75.0
octave-2$87.5
eleven_v3$100
speech-2.8-hd$100.0