Voice AI BenchmarksBy

PROVIDERLAST 30 DAYS

official resource

OpenAI voice AI models and benchmarks

OpenAI lists 6 STT, TTS and S2S models in Coval. Fastest dated mean latency over 30 days: STT: Whisper Large v3 on Baseten (dedicated inference) at 131 ms TTFS. TTS: GPT-4o mini TTS at 2585 ms TTFA. Last measured .

OpenAI creates voice models across transcription, synthesis and native speech-to-speech interaction.

Measured models
6
STTTTSS2S
Best STT model#11 / 29
131ms
Whisper Large v3
Best TTS model#32 / 32
2585ms
GPT-4o mini TTS

Overview

OpenAI ships voice models across STT, TTS and S2S — dedicated Audio API models, Realtime paths, and open-weight Whisper served by third parties.

Every model below is measured daily on the same fixed inputs and ranked against the full field, never blended into a company score.

Model lineup

Speech-to-Text
Whisper Large v3, GPT Realtime Whisper, GPT-4o Transcribe, GPT-4o mini Transcribe
Text-to-Speech
GPT-4o mini TTS
Speech-to-Speech
GPT Realtime 2

Speech-to-Text

Full STT dashboard

Ranked on Time to Final Segment against 29 measured models.

Time to Final Segmentms · lower is better · OpenAI models markedEvery measured STT model on Time to Final Segment, with OpenAI's models highlighted.
  1. #1Qwen3 ASR 1.7b39 ms
  2. #2STT RT v545 ms
  3. #3Qwen3 ASR Fast47 ms
  4. #4STT 167 ms
  5. #5Nova 386 ms
  6. #6Nova 287 ms
  7. #7Parakeet TDT 0.6B v397 ms
  8. #11Whisper Large v3via Baseten131 ms
  9. #15Whisper Large v3via Together AI198 ms
  10. #24GPT Realtime Whisper599 ms
  11. #26GPT-4o mini Transcribe710 ms
  12. #28GPT-4o Transcribe784 ms
leaders plus OpenAI models · 17 other models in the full table
Word Error Rate% · lower is better · OpenAI models markedEvery measured STT model on Word Error Rate, with OpenAI's models highlighted.
  1. #1Universal 3.6 Pro2.1%
  2. #2Universal 3.5 Pro2.4%
  3. #3Qwen3 ASR Fast2.6%
  4. #4resonant-12.6%
  5. #5Qwen3 ASR 1.7b2.9%
  6. #6Gemini 3.5 Transcribe Live3.0%
  7. #7Chirp 33.2%
  8. #9GPT-4o Transcribe3.8%
  9. #11GPT-4o mini Transcribe3.9%
  10. #13GPT Realtime Whisper4.0%
  11. #16Whisper Large v3via Baseten4.5%
  12. #27Whisper Large v3via Together AI7.4%
leaders plus OpenAI models · 19 other models in the full table
Time to First Tokenms · lower is better · OpenAI models markedEvery measured STT model on Time to First Token, with OpenAI's models highlighted.
  1. #1Whisper Large v3via Baseten924 ms
  2. #2Qwen3 ASR 1.7b946 ms
  3. #3Universal 3.5 Pro1026 ms
  4. #4Universal 3.6 Pro1071 ms
  5. #5Flux1091 ms
  6. #6Flux Multilingual1165 ms
  7. #7STT 11205 ms
  8. #9Whisper Large v3via Together AI1258 ms
  9. #19GPT Realtime Whisper1824 ms
leaders plus OpenAI models · 18 other models in the full table
TTFS vs WER — OpenAI vs every measured modelEach point is one measured STT model · OpenAI models highlighted · 30-day averagesTime to Final Segment against Word Error Rate for every measured STT model, with OpenAI's models 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 — 86 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%Universal 3.6 Pro — 135 ms, 2.1%Ink 2 — 149 ms, 3.6%Universal 3.5 Pro — 170 ms, 2.4%Pulse — 205 ms, 4.6%Grok STT — 223 ms, 4.1%resonant-1 — 252 ms, 2.6%Default — 261 ms, 8.8%Linden 1 — 274 ms, 3.9%Gemini 3.5 Transcribe Live — 317 ms, 3.0%Daytona STT — 399 ms, 4.9%Voxtral Mini Transcribe Realtime 2602 — 415 ms, 5.4%Solaria 1 — 620 ms, 6.1%Chirp 3 — 768 ms, 3.2%Chirp 2 — 909 ms, 3.9%Whisper Large v3 via Baseten — 131 ms, 4.5%Whisper Large v3 via Together AI — 198 ms, 7.4%GPT Realtime Whisper — 599 ms, 4.0%GPT-4o mini Transcribe — 710 ms, 3.9%GPT-4o Transcribe — 784 ms, 3.8%
Benchmarked models with their Time to Final Segment over the last 30 days.
ModelHostTTFSRank
Whisper Large v3Baseten131 ms11th of 29
GPT Realtime WhisperOpenAI599 ms24th of 29
GPT-4o TranscribeOpenAI784 ms28th of 29
GPT-4o mini TranscribeOpenAI710 ms26th of 29

Text-to-Speech

Full TTS dashboard

Ranked on Time to First Audio against 32 measured models.

Time to First Audioms · lower is better · OpenAI models markedEvery measured TTS model on Time to First Audio, with OpenAI's models highlighted.
  1. #1vui59 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 TTS120 ms
  8. #32GPT-4o mini TTS2585 ms
leaders plus OpenAI models · 24 other models in the full table
Word Error Rate% · lower is better · OpenAI models markedEvery measured TTS model on Word Error Rate, with OpenAI's models highlighted.
  1. #1Sonic 3.61.5%
  2. #2S2.1 Pro Free1.5%
  3. #3Simba 3.01.5%
  4. #4S2.1 Pro1.5%
  5. #5Lightning v3.1 Pro1.5%
  6. #6Sonic 3.51.5%
  7. #7Chirp 3 HD1.5%
  8. #15GPT-4o mini TTS1.8%
leaders plus OpenAI models · 24 other models in the full table
TTFA vs WER — OpenAI vs every measured modelEach point is one measured TTS model · OpenAI models highlighted · 30-day averagesTime to First Audio against Word Error Rate for every measured TTS model, with OpenAI's models highlighted.
vui — 59 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 — 120 ms, 2.6%Simba 3.2 — 126 ms, 1.9%drama3-preview — 167 ms, 2.0%TTS 2 — 191 ms, 1.6%Flash v2.5 — 192 ms, 1.8%Eleven v4 Turbo — 201 ms, 1.7%Flux TTS — 205 ms, 1.7%TTS RT v1 — 255 ms, 2.4%TTS Rt v2 — 256 ms, 2.4%Mist v3 — 287 ms, 2.1%Phantom Z 3.4 conversational — 291 ms, 1.8%Sonic 3.5 — 293 ms, 1.5%Aura 2 — 297 ms, 1.8%Sonic 3.6 — 299 ms, 1.5%Falcon 2 — 300 ms, 1.7%Coda — 332 ms, 9.5%S2.1 Pro — 341 ms, 1.5%Grok TTS — 363 ms, 1.6%S1 — 389 ms, 1.8%Lightning v3.1 Pro — 392 ms, 1.5%Daytona TTS — 486 ms, 1.6%Aura 2 En — 569 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 — 2585 ms, 1.8%
Benchmarked models with their Time to First Audio over the last 30 days.
ModelHostTTFARank
GPT-4o mini TTSOpenAI2585 ms32nd of 32

Speech-to-Speech

Full S2S dashboard

Ranked on Voice-to-Voice Latency.

Benchmarked models with their Voice-to-Voice Latency over the last 30 days.
ModelHostV2VRank
GPT Realtime 2OpenAIno measurements this window

How fast are OpenAI's STT, TTS and S2S models?

Whisper Large v3 on Baseten (dedicated inference) measures mean 131 ms time to final segment (11th of 29) among STT systems. Last measured 2026-10-14. GPT Realtime Whisper measures mean 599 ms time to final segment (24th of 29) among STT systems. Last measured 2026-10-14. GPT-4o mini Transcribe measures mean 710 ms time to final segment (26th of 29) among STT systems. Last measured 2026-10-14. GPT-4o Transcribe measures mean 784 ms time to final segment (28th of 29) among STT systems. Last measured 2026-10-14. GPT-4o mini TTS measures mean 2585 ms time to first audio (32nd of 32) among TTS systems. Last measured 2026-10-14. No dated, ranked S2S latency is available in this window.

How accurate are OpenAI's STT, TTS and S2S models?

GPT-4o Transcribe measures 3.8% word error rate (9th of 31) among STT systems. Last measured 2026-10-14. GPT-4o mini Transcribe measures 3.9% word error rate (11th of 31) among STT systems. Last measured 2026-10-14. GPT Realtime Whisper measures 4.0% word error rate (13th of 31) among STT systems. Last measured 2026-10-14. Whisper Large v3 on Baseten (dedicated inference) measures 4.5% word error rate (16th of 31) among STT systems. Last measured 2026-10-14. GPT-4o mini TTS measures 1.8% word error rate (15th of 32) among TTS systems. Last measured 2026-10-14. No dated, ranked S2S instruction adherence is available in this window.

Which OpenAI model is fastest?

Its fastest dated STT result is Whisper Large v3 on Baseten (dedicated inference) at mean 131 ms time to final segment (11th of 29) among STT systems, with 4.5% WER. Last measured 2026-10-14. Its fastest dated TTS result is GPT-4o mini TTS at mean 2585 ms time to first audio (32nd of 32) among TTS systems, with 1.8% WER. Last measured 2026-10-14.

Limits of this comparison

Coval measures OpenAI's first-party APIs; Whisper Large v3 also appears where other providers host its open weights.

  • The different API paths are measured separately and do not cover every voice or session configuration.

Official resources

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