Soniox TTS RT v1
Soniox TTS RT v1 is a proprietary text-to-speech (TTS) API from Soniox, measured by two independent benchmarks: Coval and Speko.
263 ms median and 325 ms P95 time to first audio on Coval (30 days to October 22, 2026): Visible delay speed tier.
Benchmark summary: Visible delay speed tier on Coval's P95 (the voice becomes a visible part of the delay), a tie within 10 ms with Soniox TTS RT v1 on slow turns. On the Speko speed and naturalness frontier: no other model is both faster and rated higher.
Soniox TTS Real-Time v2 is a proprietary text-to-speech (TTS) API from Soniox, measured by three independent benchmarks: Artificial Analysis, Coval and Speko.
Speed tier
Visible delayP95 of 320 ms or more. The voice becomes a visible part of the delay.
Data read October 22, 2026. No combined score: methodology
Gradium TTS vs Soniox TTS Real-Time v2 Fastest TTS APIs, by speed tierOn Coval's 30-day window ending October 22, 2026 (2,960 runs from US East), Soniox TTS Real-Time v2 starts speaking after 262.7 ms on a typical turn (median) and 324.9 ms on a slow turn (P95, the slowest one in twenty). That P95 is 15th lowest of the 32 models Coval measures, which puts it in the Visible delay tier: the voice becomes a visible part of the delay. On the median it is a tie (gap under 10 ms) with Deepgram Aura 2, Rime Mist v3 and Soniox TTS RT v1. Speko's board plots it at 134 ms time to first audio from US East; its model page reports a 381 ms sweep.
Speko rates its naturalness at 1,606 Elo from blind A/B votes on phone-agent lines (5th highest of the 43 models Speko rates). In the Artificial Analysis Speech Arena it holds 1,180 ±11 Elo (95% interval), 22nd highest of the 98 models Artificial Analysis rates; scores whose intervals overlap are a tie. In its Customer Service category, Artificial Analysis places it 22nd, at 1,209 Elo.
Speko's pronunciation test on numbers, dates and currency scores it 0.49 on a 0-to-1 scale (9th highest of the 30 models Speko scores). This is the accuracy measure used here; Coval's word error rate separates models too little to compare them.
Coval lists $14.1 per 1M characters (since Sep 22, 2026). Artificial Analysis lists $14.2. Speko lists about $13. List prices, as each source prints them: a range of $13 to $14.2.
Tested in Spanish, German, French, Japanese, Chinese, Arabic, Filipino, Norwegian and Thai by Artificial Analysis or Speko. A language missing here was not tested, which does not mean it is unsupported.
Soniox TTS RT v1 is a proprietary text-to-speech (TTS) API from Soniox, measured by two independent benchmarks: Coval and Speko.
Rime Mist v3 is a proprietary text-to-speech (TTS) API from Rime, measured by two independent benchmarks: Coval and Speko.
Cartesia Sonic 3.6 is a proprietary text-to-speech (TTS) API from Cartesia, measured by three independent benchmarks: Artificial Analysis, Coval and Speko.