SpeechifyAI Simba 3.2 Speechify · text-to-speech (TTS) API

Real-time latency

On Coval's 30-day window ending October 1, 2026 (11,833 runs), SpeechifyAI Simba 3.2 has a P50 (median) time-to-first-audio (TTFA) of 362.7 ms, 22nd lowest of the 31 models Coval measures. Its P99, the slowest 1% of turns, is 921.5 ms (26th lowest of the 31 models Coval measures). Latency consistency: a standard deviation of 388.2 ms (26th lowest of the 31 models Coval measures). Mean time to first byte (TTFB) is 298.8 ms. Speko measures a P50 TTFA of 85 ms from US East (n=30), dated July 3, 2026: 5th lowest of the 29 models Speko measures.

Accuracy (word error rate)

Coval measures an accuracy (word error rate) of 4.3% over 11,826 samples on the same window, 5th lowest of the 31 models Coval measures. Speko rates its pronunciation robustness on numbers, dates and currency at 0.21 on a 0-to-1 scale (20th highest of the 28 models Speko rates).

Voice quality

In the Artificial Analysis Speech Arena (capture of October 1, 2026), SpeechifyAI Simba 3.2 has an Elo of 1,239 (±13, 95% confidence) from 2,669 blind comparisons, 7th highest of the 92 models Artificial Analysis rates. Speko rates its naturalness at 1,573 Elo from blind A/B votes (13th highest of the 41 models Speko rates).

Price

Artificial Analysis lists SpeechifyAI Simba 3.2 at $6.6 per 1M characters (6th lowest of the 78 models Artificial Analysis lists a price for). Speko lists a cost of about $10 per 1M characters.

Scoring note

Artificial Analysis changed its Speech Arena Elo between the captures of September 30, 2026 and October 1, 2026 without a new measurement date (1,241, then 1,239): the score uses the average, 1,240. The value shown is the latest. Speko changed its pronunciation robustness between the captures of September 30, 2026 and October 1, 2026 without a new measurement date (0.7, then 0.21): the score uses the average, 0.455. The value shown is the latest.