ModelCards
← All benchmarks

Median time to first speech

We haven't written a description for this benchmark yet; see the card notes below for how each maker ran it.

Your workloadRAG over docs · 12K in · 600 out · 50K requests/mo · 30% cached
  1. 1Eleven v4 Turbo · ElevenLabs
    150 ms

“Best value”: no other model here reports a higher score at a lower cost for your workload. Scores are self-reported by each maker with different settings, so treat close calls as ties.

How each maker reported it

  • Eleven v4 Turbo launch post (footnote 3)

    ElevenLabs' own measurement (Sept 2026, network latency removed, Turbo over WebSocket streaming): median time from request to audible speech ~150 ms, against Cartesia Sonic 3.6, xAI TTS, Gemini 3.8 Flash-Lite TTS and GPT-4o mini TTS, which the chart puts at 262–814 ms. The ~75% blind-preference and Artificial Analysis #1 claims in the post refer to Eleven v4, not Turbo.

Only each model's own maker's number is shown; scores a maker reports for rival models are left out.