Voice AI BenchmarksBy

SONIOXTEXT-TO-SPEECH6,387 SAMPLES / 30 DAYSLAST RUN OCT 14, 2026, 10:30 UTC

official resource

TTS Rt v2 text-to-speech benchmarks

TTS Rt v2, hosted by Soniox, measures mean 256 ms time to first audio (15th of 32) and 2.4% word error rate (28th of 32) among TTS systems. Results cover the last 30 days. Coval benchmarks hosted endpoints daily using a consistent evaluation methodology. Last measured .

TTS RT v2 is Soniox's newer streaming synthesis endpoint.

Time to First Audio#15 / 32
256ms
TTFA Network Roundtrip#20 / 32
181ms
TTFA Leading Silence#20 / 32
75ms
Word Error Rate#28 / 32
2.4%

Overview

The official Soniox service supports multilingual synthesis and voice cloning while retaining a distinct versioned ID.

TTS Rt v2 is tested every day on a fixed set of text prompts, measuring how quickly audible speech starts and how intelligible the result is.

Technical specifications

Made by
Soniox
Hosted by
Soniox
Source
Official API
Licensing
Proprietary
Deployment
Cloud
Region
US
Features
Multilingual, Voice cloning

How TTS Rt v2 ranks

Full TTS dashboard
Time to First Audio — every measured TTS modelMilliseconds · lower is better · 30-day averageEvery TTS model ranked on Time to First Audio, with TTS Rt v2 highlighted.
  1. #1vui60 ms
  2. #2Default63 ms
  3. #3Qwen3 TTS Fast68 ms
  4. #4TTS Flash 291 ms
  5. #5Qwen3 TTS 1.7b107 ms
  6. #6Simba 3.0111 ms
  7. #7Palabra TTS v1120 ms
  8. #8Simba 3.2126 ms
  9. #9drama3-preview169 ms
  10. #10TTS 2191 ms
  11. #11Flash v2.5191 ms
  12. #12Eleven v4 Turbo201 ms
  13. #13Flux TTS205 ms
  14. #14TTS RT v1254 ms
  15. #15TTS Rt v2256 ms
Show all 32 models
  1. #16Mist v3287 ms
  2. #17Sonic 3.5293 ms
  3. #18Phantom Z 3.4 conversational293 ms
  4. #19Aura 2297 ms
  5. #20Sonic 3.6299 ms
  6. #21Falcon 2301 ms
  7. #22Coda332 ms
  8. #23S2.1 Pro341 ms
  9. #24Grok TTS362 ms
  10. #25S1389 ms
  11. #26Lightning v3.1 Pro393 ms
  12. #27Daytona TTS486 ms
  13. #28Aura 2 En570 ms
  14. #29Chirp 3 HD597 ms
  15. #30Qwen3 TTS Flash Realtime718 ms
  16. #31S2.1 Pro Free850 ms
  17. #32GPT-4o mini TTS2581 ms
median of all models · 290 ms
TTS models over the last 30 days, ranked on Time to First Audio.
#ModelHostTTFAWERSamples
1vuiFluxions60 ms1,724
2Gradium TTS Beta 202609Gradium63 ms1,896
3Qwen3 TTS FastNari Labs68 ms1,896
4TTS Flash 2Inworld AI91 ms1,700
5Qwen3 TTS 1.7bBaseten107 ms1,248
6Simba 3.0SpeechifyAI111 ms1,895
7Palabra TTS v1Palabra120 ms1,558
8Simba 3.2SpeechifyAI126 ms1,892
9Drama 3 Preview Us EastFish Audio169 ms142
10TTS 2Inworld AI191 ms1,700
11Flash v2.5ElevenLabs191 ms1,887
12Eleven v4 TurboElevenLabs201 ms1,206
13Flux TTSDeepgram205 ms1,759
14TTS RT v1Soniox254 ms1,599
15TTS Rt v2Soniox256 ms1,597
16Mist v3Rime287 ms602
17Sonic 3.5Cartesia293 ms1,896
18Phantom Z 3.4 conversationalDeepdub293 ms1,631
19Aura 2Deepgram297 ms1,828
20Sonic 3.6Cartesia299 ms1,896
21Falcon 2Murf301 ms1,895
22CodaRime332 ms602
23S2.1 ProFish Audio341 ms1,896
24Grok TTSSpaceXAI362 ms1,891
25S1Fish Audio389 ms1,895
26Lightning v3.1 ProSmallest393 ms1,896
27Daytona TTSGuava486 ms1,152
28Aura 2 EnCloudflare570 ms1,895
29Chirp 3 HDGoogle597 ms1,896
30Qwen3 TTS Flash RealtimeAlibaba718 ms1,895
31S2.1 Pro FreeFish Audio850 ms1,889
32GPT-4o mini TTSOpenAI2581 ms1,832
Under-sampled models are excluded; tied models share a place. Dotted WER values split into substitutions, deletions and insertions on hover or tap.

Latency vs accuracy

TTFA vs WEREach point is one measured model · TTS Rt v2 highlighted · 30-day averagesTime to First Audio against Word Error Rate for every measured TTS model, with TTS Rt v2 highlighted.
vui — 60 ms, 2.1%Default — 63 ms, 1.9%Qwen3 TTS Fast — 68 ms, 1.8%TTS Flash 2 — 91 ms, 1.7%Qwen3 TTS 1.7b — 107 ms, 1.8%Simba 3.0 — 111 ms, 1.5%Palabra TTS v1 — 120 ms, 2.6%Simba 3.2 — 126 ms, 1.9%drama3-preview — 169 ms, 2.0%TTS 2 — 191 ms, 1.6%Flash v2.5 — 191 ms, 1.8%Eleven v4 Turbo — 201 ms, 1.7%Flux TTS — 205 ms, 1.7%TTS RT v1 — 254 ms, 2.4%Mist v3 — 287 ms, 2.1%Sonic 3.5 — 293 ms, 1.5%Phantom Z 3.4 conversational — 293 ms, 1.8%Aura 2 — 297 ms, 1.8%Sonic 3.6 — 299 ms, 1.5%Falcon 2 — 301 ms, 1.7%Coda — 332 ms, 9.5%S2.1 Pro — 341 ms, 1.5%Grok TTS — 362 ms, 1.6%S1 — 389 ms, 1.8%Lightning v3.1 Pro — 393 ms, 1.5%Daytona TTS — 486 ms, 1.6%Aura 2 En — 570 ms, 2.1%Chirp 3 HD — 597 ms, 1.5%Qwen3 TTS Flash Realtime — 718 ms, 3.0%S2.1 Pro Free — 850 ms, 1.5%GPT-4o mini TTS — 2581 ms, 1.8%TTS Rt v2 — 256 ms, 2.4%

Where the errors come from

WER compositionTTS Rt v2's Word Error Rate split by error type · 30-day averageTTS Rt v2's WER split into substitutions, deletions and insertions.
  • TTS Rt v22.4%
SubstitutionsDeletionsInsertions

Averages and tail latency

Averages hide slow outliers — these are the distributions behind each figure.

TTS Rt v2 latency distributionMilliseconds · shared axis across metrics · last 30 daysTTS Rt v2's latency percentiles per metric: p25–p75 band, p50 tick, whisker to max.
  • Time to First Audiop50 259 ms · Max 739 ms
  • TTFA Network Roundtripp50 190 ms · Max 654 ms
  • TTFA Leading Silencep50 75 ms · Max 120 ms
band p25–p75 · tick p50 · whisker to max with p90 and p95 stops
Average and percentile values per metric, with the number of samples behind each row.
MetricAveragep25p50p75p90p95MaxSamples
Time to First Audio256 ms217 ms259 ms290 ms307 ms319 ms739 ms1,597
TTFA Network Roundtrip181 ms142 ms190 ms212 ms228 ms237 ms654 ms1,597
TTFA Leading Silence75 ms65 ms75 ms86 ms92 ms94 ms120 ms1,597
Word Error Rate2.4%0.0%0.0%2.8%9.3%15.2%43.8%1,596

Last 30 days

Daily medians from the same measurement runs · gaps are days without qualifying runs.

Time to First Audio — daily p50Line p50 · band p25–p75 · UTC daysTTS Rt v2's daily median Time to First Audio over the last 30 days.
TTS Rt v2 · Sep 23: 310 msTTS Rt v2 · Sep 24: 233 msTTS Rt v2 · Sep 25: 278 msTTS Rt v2 · Sep 26: 224 msTTS Rt v2 · Sep 27: 289 msTTS Rt v2 · Sep 28: 209 msTTS Rt v2 · Sep 29: 293 msTTS Rt v2 · Sep 30: 237 msTTS Rt v2 · Oct 1: 295 msTTS Rt v2 · Oct 2: 245 msTTS Rt v2 · Oct 3: 245 msTTS Rt v2 · Oct 4: 292 msTTS Rt v2 · Oct 5: 247 msTTS Rt v2 · Oct 6: 236 msTTS Rt v2 · Oct 7: 269 msTTS Rt v2 · Oct 8: 260 msTTS Rt v2 · Oct 9: 288 msTTS Rt v2 · Oct 10: 219 msTTS Rt v2 · Oct 11: 257 msTTS Rt v2 · Oct 12: 269 msTTS Rt v2 · Oct 13: 264 msTTS Rt v2 · Oct 14: 252 ms
Word Error Rate — daily averageDaily average · UTC daysTTS Rt v2's daily Word Error Rate over the last 30 days.
TTS Rt v2 · Sep 23: 3.1%TTS Rt v2 · Sep 24: 3.0%TTS Rt v2 · Sep 25: 5.0%TTS Rt v2 · Sep 26: 4.0%TTS Rt v2 · Sep 27: 7.0%TTS Rt v2 · Sep 28: 5.9%TTS Rt v2 · Sep 29: 4.3%TTS Rt v2 · Sep 30: 4.1%TTS Rt v2 · Oct 1: 7.9%TTS Rt v2 · Oct 2: 3.0%TTS Rt v2 · Oct 3: 7.1%TTS Rt v2 · Oct 4: 9.5%TTS Rt v2 · Oct 5: 4.7%TTS Rt v2 · Oct 6: 5.3%TTS Rt v2 · Oct 7: 5.4%TTS Rt v2 · Oct 8: 6.3%TTS Rt v2 · Oct 9: 3.0%TTS Rt v2 · Oct 10: 3.2%TTS Rt v2 · Oct 11: 4.5%TTS Rt v2 · Oct 12: 4.4%TTS Rt v2 · Oct 13: 6.1%TTS Rt v2 · Oct 14: 7.2%

How fast is TTS Rt v2?

On Soniox, TTS Rt v2 measures mean 256 ms time to first audio (15th of 32). Last measured 2026-10-14.

How accurate is TTS Rt v2?

On Soniox, TTS Rt v2 measures 2.4% word error rate (28th of 32). Last measured 2026-10-14.

Who hosts TTS Rt v2?

TTS Rt v2 is created by Soniox and served by Soniox. Coval measures each hosted endpoint separately.

Limits of this comparison

RT v2 receives the same prompt set as v1 and other measured TTS models.

  • A version comparison covers measured TTS metrics, not preference, expressiveness or clone similarity.

Official sources

Results are re-measured daily using fixed datasets and reported over a rolling 30-day window. Same datasets, prompts and metric definitions for every model, measured by Coval’s open-source runner. Full methodology on the overview.

Evaluate your own voice agent

Use Coval to test your production configuration, prompts and calls—not only the public benchmark endpoints.

Book a Demo