Voice AI BenchmarksBy

OPENAISPEECH-TO-TEXT133,110 SAMPLES / 30 DAYSLAST RUN OCT 14, 2026, 11:00 UTC

official resource

GPT Realtime Whisper speech-to-text benchmarks

GPT Realtime Whisper, hosted by OpenAI, measures mean 599 ms time to final segment (24th of 29) and 4.0% word error rate (13th of 31) among STT systems. Results cover the last 30 days. Coval benchmarks hosted endpoints daily using a consistent evaluation methodology. Last measured .

GPT Realtime Whisper is Whisper transcription running inside an OpenAI Realtime session, rather than the standalone Audio API.

Time to Final Segment#24 / 29
599ms
Word Error Rate#13 / 31
4%
Time to First Token#19 / 27
1824ms

Overview

The path runs within a persistent Realtime connection while still being scored against Coval's standard STT references.

GPT Realtime Whisper is tested every day on fixed public audio — clean, accented, noisy, reverberant, far-field, clipped and phone-codec speech — for transcription accuracy and streaming latency.

Technical specifications

Made by
OpenAI
Hosted by
OpenAI
Source
Official API
Licensing
Proprietary
Deployment
Cloud
Region
US
Features
Multilingual

How GPT Realtime Whisper ranks

Full STT dashboard
Time to Final Segment — every measured STT modelMilliseconds · lower is better · 30-day averageEvery STT model ranked on Time to Final Segment, with GPT Realtime Whisper highlighted.
  1. #1Qwen3 ASR 1.7b39 ms
  2. #2STT RT v545 ms
  3. #3Qwen3 ASR Fast47 ms
  4. #4STT 167 ms
  5. #5Nova 385 ms
  6. #6Nova 287 ms
  7. #7Parakeet TDT 0.6B v397 ms
  8. #8Flux Multilingual98 ms
  9. #9Flux99 ms
  10. #10Scribe v2 Realtime122 ms
  11. #11Whisper Large v3via Baseten131 ms
  12. #12Universal 3.6 Pro135 ms
  13. #13Ink 2149 ms
  14. #14Universal 3.5 Pro168 ms
  15. #15Whisper Large v3via Together AI199 ms
  16. #16Pulse205 ms
  17. #17Grok STT223 ms
  18. #18resonant-1252 ms
  19. #19Default258 ms
  20. #20Linden 1271 ms
  21. #21Gemini 3.5 Transcribe Live317 ms
  22. #22Daytona STT399 ms
  23. #23Voxtral Mini Transcribe Realtime 2602411 ms
  24. #24GPT Realtime Whisper599 ms
Show all 29 models
  1. #25Solaria 1626 ms
  2. #26GPT-4o mini Transcribe708 ms
  3. #27Chirp 3768 ms
  4. #28GPT-4o Transcribe783 ms
  5. #29Chirp 2908 ms
median of all models · 199 ms
STT models over the last 30 days, ranked on Time to Final Segment.
#ModelHostTTFSWERTTFTSamples
1Qwen3 ASR 1.7bBaseten39 ms946 ms13,669
2STT RT v5Soniox45 ms1518 ms23,533
3Qwen3 ASR FastNari Labs47 ms1752 ms26,653
4STT 1Inworld AI67 ms1208 ms24,749
5Nova 3Deepgram85 ms1432 ms26,002
6Nova 2Deepgram87 ms1417 ms25,960
7Parakeet TDT 0.6B v3Together AI97 ms1228 ms26,462
8Flux MultilingualDeepgram98 ms1165 ms26,007
9FluxDeepgram99 ms1091 ms26,005
10Scribe v2 RealtimeElevenLabs122 ms2157 ms26,640
11Whisper Large v3Baseten131 ms924 ms13,663
12Universal 3.6 ProAssemblyAI135 ms1072 ms24,394
13Ink 2Cartesia149 ms1967 ms26,653
14Universal 3.5 ProAssemblyAI168 ms1026 ms26,654
15Whisper Large v3Together AI199 ms1260 ms26,447
16PulseSmallest205 ms1976 ms26,654
17Grok STTSpaceXAI223 ms—26,576
18resonant-1Reson8252 ms—26,654
19DefaultGradium258 ms1986 ms26,632
20Linden 1Speechmatics271 ms1537 ms26,595
21Gemini 3.5 Transcribe LiveGemini317 ms1606 ms26,567
22Daytona STTGuava399 ms5695 ms5,225
23Voxtral Mini Transcribe Realtime 2602Mistral411 ms1851 ms26,158
24GPT Realtime WhisperOpenAI599 ms1824 ms26,622
25Solaria 1Gladia626 ms1775 ms26,215
26GPT-4o mini TranscribeOpenAI708 ms—26,621
27Chirp 3Google768 ms5956 ms26,638
28GPT-4o TranscribeOpenAI783 ms—26,625
29Chirp 2Google908 ms6099 ms26,650
—Nemotron 3.5 ASR StreamingTogether AI—1544 ms26,550
—Universal StreamingAssemblyAI—1509 ms26,625
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: 13th of 31 on Word Error Rate.

Latency vs accuracy

TTFS vs WEREach point is one measured model · GPT Realtime Whisper highlighted · 30-day averagesTime to Final Segment against Word Error Rate for every measured STT model, with GPT Realtime Whisper highlighted.
Qwen3 ASR 1.7b — 39 ms, 2.9%STT RT v5 — 45 ms, 4.8%Qwen3 ASR Fast — 47 ms, 2.6%STT 1 — 67 ms, 7.1%Nova 3 — 85 ms, 5.3%Nova 2 — 87 ms, 6.7%Parakeet TDT 0.6B v3 — 97 ms, 11.1%Flux Multilingual — 98 ms, 7.5%Flux — 99 ms, 5.8%Scribe v2 Realtime — 122 ms, 4.4%Whisper Large v3 via Baseten — 131 ms, 4.5%Universal 3.6 Pro — 135 ms, 2.1%Ink 2 — 149 ms, 3.6%Universal 3.5 Pro — 168 ms, 2.3%Whisper Large v3 via Together AI — 199 ms, 7.4%Pulse — 205 ms, 4.6%Grok STT — 223 ms, 4.1%resonant-1 — 252 ms, 2.6%Default — 258 ms, 8.8%Linden 1 — 271 ms, 3.9%Gemini 3.5 Transcribe Live — 317 ms, 3.0%Daytona STT — 399 ms, 4.9%Voxtral Mini Transcribe Realtime 2602 — 411 ms, 5.3%Solaria 1 — 626 ms, 6.0%GPT-4o mini Transcribe — 708 ms, 3.9%Chirp 3 — 768 ms, 3.2%GPT-4o Transcribe — 783 ms, 3.7%Chirp 2 — 908 ms, 3.9%GPT Realtime Whisper — 599 ms, 4.0%

Where the errors come from

WER compositionGPT Realtime Whisper's Word Error Rate split by error type · 30-day averageGPT Realtime Whisper's WER split into substitutions, deletions and insertions.
  • GPT Realtime Whisper4.0%
SubstitutionsDeletionsInsertions

Averages and tail latency

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

GPT Realtime Whisper latency distributionMilliseconds · shared axis across metrics · last 30 daysGPT Realtime Whisper's latency percentiles per metric: p25–p75 band, p50 tick, whisker to max.
  • Time to Final Segmentp50 554 ms · Max 25534 ms
  • Time to First Tokenp50 1880 ms · Max 12852 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 Final Segment599 ms490 ms554 ms606 ms653 ms707 ms25534 ms26,622
Word Error Rate4.0%0.0%0.0%6.3%13.3%20.0%100.0%26,622
Time to First Token1824 ms1521 ms1880 ms2079 ms2321 ms2638 ms12852 ms26,622

Last 30 days

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

Time to Final Segment — daily p50Line p50 · band p25–p75 · UTC daysGPT Realtime Whisper's daily median Time to Final Segment over the last 30 days.
GPT Realtime Whisper · Sep 14: 521 msGPT Realtime Whisper · Sep 15: 543 msGPT Realtime Whisper · Sep 16: 532 msGPT Realtime Whisper · Sep 17: 552 msGPT Realtime Whisper · Sep 18: 566 msGPT Realtime Whisper · Sep 19: 589 msGPT Realtime Whisper · Sep 20: 535 msGPT Realtime Whisper · Sep 21: 569 msGPT Realtime Whisper · Sep 22: 555 msGPT Realtime Whisper · Sep 23: 565 msGPT Realtime Whisper · Sep 24: 579 msGPT Realtime Whisper · Sep 25: 541 msGPT Realtime Whisper · Sep 26: 521 msGPT Realtime Whisper · Sep 27: 503 msGPT Realtime Whisper · Sep 28: 545 msGPT Realtime Whisper · Sep 29: 555 msGPT Realtime Whisper · Sep 30: 547 msGPT Realtime Whisper · Oct 1: 523 msGPT Realtime Whisper · Oct 2: 575 msGPT Realtime Whisper · Oct 3: 568 msGPT Realtime Whisper · Oct 4: 545 msGPT Realtime Whisper · Oct 5: 562 msGPT Realtime Whisper · Oct 6: 565 msGPT Realtime Whisper · Oct 7: 519 msGPT Realtime Whisper · Oct 8: 553 msGPT Realtime Whisper · Oct 9: 543 msGPT Realtime Whisper · Oct 10: 571 msGPT Realtime Whisper · Oct 11: 560 msGPT Realtime Whisper · Oct 12: 530 msGPT Realtime Whisper · Oct 13: 535 msGPT Realtime Whisper · Oct 14: 551 ms
Word Error Rate — daily averageDaily average · UTC daysGPT Realtime Whisper's daily Word Error Rate over the last 30 days.
GPT Realtime Whisper · Sep 14: 4.9%GPT Realtime Whisper · Sep 15: 4.4%GPT Realtime Whisper · Sep 16: 5.0%GPT Realtime Whisper · Sep 17: 5.5%GPT Realtime Whisper · Sep 18: 5.0%GPT Realtime Whisper · Sep 19: 5.2%GPT Realtime Whisper · Sep 20: 5.6%GPT Realtime Whisper · Sep 21: 4.4%GPT Realtime Whisper · Sep 22: 6.1%GPT Realtime Whisper · Sep 23: 4.6%GPT Realtime Whisper · Sep 24: 5.6%GPT Realtime Whisper · Sep 25: 6.1%GPT Realtime Whisper · Sep 26: 5.9%GPT Realtime Whisper · Sep 27: 5.9%GPT Realtime Whisper · Sep 28: 4.5%GPT Realtime Whisper · Sep 29: 4.3%GPT Realtime Whisper · Sep 30: 5.8%GPT Realtime Whisper · Oct 1: 4.3%GPT Realtime Whisper · Oct 2: 4.7%GPT Realtime Whisper · Oct 3: 5.8%GPT Realtime Whisper · Oct 4: 5.3%GPT Realtime Whisper · Oct 5: 5.3%GPT Realtime Whisper · Oct 6: 7.9%GPT Realtime Whisper · Oct 7: 4.2%GPT Realtime Whisper · Oct 8: 5.6%GPT Realtime Whisper · Oct 9: 4.3%GPT Realtime Whisper · Oct 10: 4.9%GPT Realtime Whisper · Oct 11: 4.6%GPT Realtime Whisper · Oct 12: 5.1%GPT Realtime Whisper · Oct 13: 4.9%GPT Realtime Whisper · Oct 14: 5.9%

Time to Final Segment by dataset

GPT Realtime Whisper by test conditionMilliseconds · lower is better · best condition firstGPT Realtime Whisper's Time to Final Segment on each benchmark dataset.
  1. WildASR clean587 ms
  2. PipeCat (production)592 ms
  3. WildASR far-field600 ms
  4. WildASR noise gaps602 ms
  5. WildASR phone codec631 ms
  6. WildASR accents637 ms
  7. WildASR clipping638 ms
Time to Final Segment per dataset, with the number of samples behind each figure.
DatasetTTFSSamples
ProductionPipeCat592 ms14,004
AccentsWildASR637 ms1,401
CleanWildASR587 ms5,607
ClippingWildASR638 ms1,402
Far-fieldWildASR600 ms1,404
Noise gapsWildASR602 ms1,401
Phone codecWildASR631 ms1,403

Strongest condition: WildASR clean at 587 ms · weakest: WildASR clipping at 638 ms.

How fast is GPT Realtime Whisper?

On OpenAI, GPT Realtime Whisper measures mean 599 ms time to final segment (24th of 29) and mean 1824 ms time to first token (19th of 27). Last measured 2026-10-14.

How accurate is GPT Realtime Whisper?

On OpenAI, GPT Realtime Whisper measures 4.0% word error rate (13th of 31). Last measured 2026-10-14.

Who hosts GPT Realtime Whisper?

GPT Realtime Whisper is created by OpenAI and served by OpenAI. Coval measures each hosted endpoint separately.

Limits of this comparison

This path is reported separately because session events and integration behavior differ from standalone audio-transcription requests.

  • The name describes Coval's integration path, and OpenAI can update Realtime defaults independently of standalone Whisper endpoints.

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