Voice AI BenchmarksBy

OPENAITEXT-TO-SPEECH45,482 SAMPLES / 30 DAYSLAST RUN SEP 15, 2026, 07:00 UTC

official resource

GPT-4o mini TTS text-to-speech benchmarks

GPT-4o mini TTS, hosted by OpenAI, measures mean 1013 ms time to first audio (28th of 28) and 4.8% word error rate (10th of 28) among TTS systems. Results cover the last 30 days. Coval benchmarks hosted endpoints daily using a consistent evaluation methodology. Last measured .

GPT-4o mini TTS is OpenAI's synthesis model with natural-language speaking-style instructions.

Time to First Audio#28 / 28
1013ms
TTFA Network Roundtrip#27 / 28
885ms
TTFA Leading Silence#22 / 28
129ms
Word Error Rate#10 / 28
4.8%

Overview

OpenAI's Speech API accepts text plus style instructions, and Coval counts silence before the first audible sample in TTFA.

GPT-4o mini TTS 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
OpenAI
Hosted by
OpenAI
Source
Official API
Licensing
Proprietary
Deployment
Cloud
Region
US
Features
Emotion control, Multilingual

How GPT-4o mini TTS 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 GPT-4o mini TTS highlighted.
  1. #1vui66 ms
  2. #2Qwen3 TTS Fast71 ms
  3. #3TTS Flash 291 ms
  4. #4Qwen3 TTS 1.7b106 ms
  5. #5Palabra TTS v1115 ms
  6. #6TTS 2181 ms
  7. #7Flash v2.5194 ms
  8. #8Default235 ms
  9. #9TTS Rt v2245 ms
  10. #10TTS RT v1245 ms
  11. #11Mist v3258 ms
  12. #12Sonic 3.5276 ms
  13. #13Phantom Z 3.4 conversational286 ms
  14. #14Aura 2308 ms
  15. #15Coda312 ms
  16. #16S2.1 Pro335 ms
  17. #17Eleven v3 Conversational348 ms
  18. #18Lightning v3.1 Pro362 ms
  19. #19S1379 ms
  20. #20Grok TTS397 ms
  21. #21Sonic 3.6423 ms
  22. #22Simba 3.2452 ms
  23. #23Simba 3.0471 ms
  24. #24Chirp 3 HD535 ms
  25. #25Falcon 2545 ms
  26. #26Qwen3 TTS Flash Realtime754 ms
  27. #27S2.1 Pro Free965 ms
  28. #28GPT-4o mini TTS1013 ms
median of all models · 310 ms
TTS models over the last 30 days, ranked on Time to First Audio.
#ModelHostTTFAWERSamples
1vuiFluxions66 ms9,450
2Qwen3 TTS FastNari71 ms2,190
3TTS Flash 2Inworld AI91 ms11,401
4Qwen3 TTS 1.7bBaseten106 ms420
5Palabra TTS v1Palabra115 ms11,385
6TTS 2Inworld AI181 ms11,401
7Flash v2.5ElevenLabs194 ms9,118
8DefaultGradium235 ms9,127
9TTS Rt v2Soniox245 ms11,232
10TTS RT v1Soniox245 ms11,234
11Mist v3Rime258 ms11,387
12Sonic 3.5Cartesia276 ms11,400
13Phantom Z 3.4 conversationalDeepdub286 ms11,396
14Aura 2Deepgram308 ms11,386
15CodaRime312 ms11,392
16S2.1 ProFish Audio335 ms11,400
17Eleven v3 ConversationalElevenLabs348 ms11,389
18Lightning v3.1 ProSmallest362 ms11,401
19S1Fish Audio379 ms11,393
20Grok TTSxAI397 ms11,386
21Sonic 3.6Cartesia423 ms8,210
22Simba 3.2Speechify452 ms11,393
23Simba 3.0Speechify471 ms11,396
24Chirp 3 HDGoogle535 ms11,400
25Falcon 2Murf545 ms11,397
26Qwen3 TTS Flash RealtimeAlibaba754 ms11,226
27S2.1 Pro FreeFish Audio965 ms11,366
28GPT-4o mini TTSOpenAI1013 ms11,370
Under-sampled models are excluded; tied models share a place. Dotted WER values split into substitutions, deletions and insertions on hover or tap.

Highest relative placement: 10th of 28 on Word Error Rate.

Latency vs accuracy

TTFA vs WEREach point is one measured model · GPT-4o mini TTS highlighted · 30-day averagesTime to First Audio against Word Error Rate for every measured TTS model, with GPT-4o mini TTS highlighted.
vui — 66 ms, 5.5%Qwen3 TTS Fast — 71 ms, 4.2%TTS Flash 2 — 91 ms, 5.4%Qwen3 TTS 1.7b — 106 ms, 5.5%Palabra TTS v1 — 115 ms, 5.8%TTS 2 — 181 ms, 4.6%Flash v2.5 — 194 ms, 6.7%Default — 235 ms, 5.3%TTS Rt v2 — 245 ms, 4.2%TTS RT v1 — 245 ms, 3.9%Mist v3 — 258 ms, 5.6%Sonic 3.5 — 276 ms, 5.9%Phantom Z 3.4 conversational — 286 ms, 6.0%Aura 2 — 308 ms, 5.2%Coda — 312 ms, 5.0%S2.1 Pro — 335 ms, 4.8%Eleven v3 Conversational — 348 ms, 4.4%Lightning v3.1 Pro — 362 ms, 4.4%S1 — 379 ms, 4.9%Grok TTS — 397 ms, 4.9%Sonic 3.6 — 423 ms, 5.3%Simba 3.2 — 452 ms, 4.4%Simba 3.0 — 471 ms, 5.2%Chirp 3 HD — 535 ms, 5.2%Falcon 2 — 545 ms, 5.2%Qwen3 TTS Flash Realtime — 754 ms, 8.9%S2.1 Pro Free — 965 ms, 4.7%GPT-4o mini TTS — 1013 ms, 4.8%

Where the errors come from

WER compositionGPT-4o mini TTS's Word Error Rate split by error type · 30-day averageGPT-4o mini TTS's WER split into substitutions, deletions and insertions.
  • GPT-4o mini TTS4.8%
SubstitutionsDeletionsInsertions

Averages and tail latency

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

GPT-4o mini TTS latency distributionMilliseconds · shared axis across metrics · last 30 daysGPT-4o mini TTS's latency percentiles per metric: p25–p75 band, p50 tick, whisker to p99.
  • Time to First Audiop50 586 ms · p99 5230 ms
  • TTFA Network Roundtripp50 493 ms · p99 5041 ms
  • TTFA Leading Silencep50 33 ms · p99 752 ms
band p25–p75 · tick p50 · whisker to p99 with p90 and p95 stops
Average and percentile values per metric, with the number of samples behind each row.
MetricAveragep25p50p75p90p95p99Samples
Time to First Audio1013 ms460 ms586 ms901 ms2797 ms3909 ms5230 ms11,370
TTFA Network Roundtrip885 ms382 ms493 ms659 ms2677 ms3806 ms5041 ms11,370
TTFA Leading Silence129 ms17 ms33 ms130 ms453 ms541 ms752 ms11,370
Word Error Rate4.8%0.0%0.0%6.3%20.0%21.4%35.7%11,372

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 daysGPT-4o mini TTS's daily median Time to First Audio over the last 30 days.
GPT-4o mini TTS · Sep 1: 516 msGPT-4o mini TTS · Sep 2: 586 msGPT-4o mini TTS · Sep 3: 522 msGPT-4o mini TTS · Sep 4: 601 msGPT-4o mini TTS · Sep 5: 569 msGPT-4o mini TTS · Sep 6: 532 msGPT-4o mini TTS · Sep 7: 563 msGPT-4o mini TTS · Sep 8: 630 msGPT-4o mini TTS · Sep 9: 555 msGPT-4o mini TTS · Sep 10: 555 msGPT-4o mini TTS · Sep 11: 542 msGPT-4o mini TTS · Sep 12: 549 msGPT-4o mini TTS · Sep 13: 574 msGPT-4o mini TTS · Sep 14: 587 msGPT-4o mini TTS · Sep 15: 575 ms
Word Error Rate — daily averageDaily average · UTC daysGPT-4o mini TTS's daily Word Error Rate over the last 30 days.
GPT-4o mini TTS · Sep 1: 4.2%GPT-4o mini TTS · Sep 2: 4.1%GPT-4o mini TTS · Sep 3: 4.6%GPT-4o mini TTS · Sep 4: 4.7%GPT-4o mini TTS · Sep 5: 4.9%GPT-4o mini TTS · Sep 6: 5.6%GPT-4o mini TTS · Sep 7: 4.7%GPT-4o mini TTS · Sep 8: 4.7%GPT-4o mini TTS · Sep 9: 6.2%GPT-4o mini TTS · Sep 10: 5.3%GPT-4o mini TTS · Sep 11: 5.0%GPT-4o mini TTS · Sep 12: 4.8%GPT-4o mini TTS · Sep 13: 4.9%GPT-4o mini TTS · Sep 14: 4.9%GPT-4o mini TTS · Sep 15: 3.7%

Time to First Audio by dataset

GPT-4o mini TTS by test conditionMilliseconds · lower is better · best condition firstGPT-4o mini TTS's Time to First Audio on each benchmark dataset.
  1. Text prompts1013 ms
Time to First Audio per dataset, with the number of samples behind each figure.
DatasetTTFASamples
Text prompts1013 ms11,370

Strongest condition: Text prompts at 1013 ms · weakest: Text prompts at 1013 ms.

How fast is GPT-4o mini TTS?

On OpenAI, GPT-4o mini TTS measures mean 1013 ms time to first audio (28th of 28). Last measured 2026-09-15.

How accurate is GPT-4o mini TTS?

On OpenAI, GPT-4o mini TTS measures 4.8% word error rate (10th of 28). Last measured 2026-09-15.

Who hosts GPT-4o mini TTS?

GPT-4o mini TTS is created by OpenAI and served by OpenAI. Coval measures each hosted endpoint separately.

Limits of this comparison

Coval measures a fixed voice and prompt configuration so changing content does not obscure API timing.

  • A single voice and instruction setting cannot represent the model's complete expressive range.

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