Humanness Index™ · TTS model
by ElevenLabs
Flash v2.5 is the multilingual member of ElevenLabs' fastest family, generating speech in about 75 ms across 32 languages.
Standings as of Jun 13, 2026, 00:15 UTC
A real arena clip: a cloned source voice reading a customer support prompt at phone quality.
Flash v2.5 is the multilingual member of ElevenLabs' fastest family, generating speech in about 75 ms across 32 languages. Announced in December 2024, it is the model ElevenLabs recommends for real time agents beyond English, and it bills at half the credit cost per character of the flagship v3. That mix makes it the default choice for latency sensitive, cost sensitive voice deployments on the ElevenLabs platform.
Sources: elevenlabs.io
Flash v2.5 pairs the Flash latency tier with 32 languages. In our 50 trial streaming benchmark it returned first audio in a median of 197 ms, the fastest ElevenLabs result on the Index.
Sources: elevenlabs.io
Standings as of Jun 13, 2026, 00:15 UTC
| Rank | Provider | Model | Humanness | Latency |
|---|---|---|---|---|
| #5 | ElevenLabs | Eleven v3 | 76 | 758 ms |
| #6 | ElevenLabs | Turbo v2.5 | 75 | 265 ms |
| #7 | ElevenLabs | Flash v2.5 | 72 | 197 ms |
| #8 | MiniMax | Speech 2.5 | 71 | 325 ms |
| #9 | Inworld | TTS-2 | 66 | 288 ms |
The benchmark is open source. Suggest a model, read the methodology, or ask us to put your voice in the arena.
