Deepgram Flux TTS Deepgram · text-to-speech (TTS) API

Speed

On Coval's 30-day window ending October 22, 2026 (3,199 runs from US East), Deepgram Flux TTS starts speaking after 199.8 ms on a typical turn (median) and 262.9 ms on a slow turn (P95, the slowest one in twenty). That P95 is 13th lowest of the 32 models Coval measures, which puts it in the Fast enough tier: fast enough for most agents. On the median it is a tie (gap under 10 ms) with ElevenLabs Eleven v4 Turbo. Speko's board plots it at 118 ms time to first audio from US East; its model page reports a 106 ms sweep.

Voice quality

Speko rates its naturalness at 1,550 Elo from blind A/B votes on phone-agent lines (17th highest of the 43 models Speko rates). In the Artificial Analysis Speech Arena it holds 1,052 ±13 Elo (95% interval), 65th highest of the 98 models Artificial Analysis rates; scores whose intervals overlap are a tie. In its Customer Service category, Artificial Analysis places it 35th, at 1,143 Elo.

Numbers and dates

Speko's pronunciation test on numbers, dates and currency scores it 0.2 on a 0-to-1 scale (25th 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.

Price

Artificial Analysis lists $45. Speko lists about $45. List prices, as each source prints them.