Official provider results
Realtime TTS-2 launch evaluation
Inworld’s launch post reports an independent speech-arena ranking, time-to-first-audio, and language coverage for Realtime TTS-2.
Source snapshots refreshed
Five measurement lanesVOICE / S2S
Reasoning, task completion, conversation dynamics, experience, and latency stay separate.
Benchmark profile
- Scope
- independent speech-arena rank, P90 server-side time-to-first-byte, and language count
- Primary metric
- Arena rank and latency
- Owner
- Inworld
- Available evidence
- Rank and latency published in the official launch post and docs; no Elo value given
What it measures
Voice experience
Is the exchange natural, robust, and responsive?
Latency
How long do responses, tool calls, and full tasks take?
Available results
- Independent speech arena
- #1
- Time to first byte (P90, server-side)
- 100 ms
- Languages
- 200+
Provider-cited at launch
Google #2, ElevenLabs #3, OpenAI #5 in Inworld’s citation.
Documented
Sub-200 ms median for the TTS layer alone.
Documented
Mid-utterance language switching with one voice identity.
Interpretation limit
Provider-reported; Inworld cites the Speech Arena rank without the underlying Elo.