Voice AI BenchmarksBy

PROVIDERLAST 30 DAYS

official resource

Google voice AI models and benchmarks

Google creates and hosts voice models across all three benchmark categories: Cloud Speech-to-Text, Cloud Text-to-Speech and the native-audio Gemini Live API.

Measured models
4
STTTTSS2S
Best STT model#23 / 24
811ms
Chirp 2
Best TTS model#24 / 30
512ms
Chirp 3 HD
Best S2S model#2 / 2
1380ms
Gemini 3.1 Flash Live (Preview)

Overview

Google ships voice models across STT, TTS and S2S; each product and generation is ranked separately in its own category.

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
Chirp 2, Chirp 3
Text-to-Speech
Chirp 3 HD
Speech-to-Speech
Gemini 3.1 Flash Live (Preview)

Speech-to-Text

Full STT dashboard

Ranked on Time to Final Segment against 24 measured models.

Time to Final Segmentms · lower is better · Google models markedEvery measured STT model on Time to Final Segment, with Google's models highlighted.
  1. #1STT RT v564 ms
  2. #2Parakeet TDT 0.6B v370 ms
  3. #3STT 183 ms
  4. #4Nova 399 ms
  5. #5Nova 2101 ms
  6. #6Ink 2108 ms
  7. #7Scribe v2 Realtime120 ms
  8. #23Chirp 2811 ms
  9. #24Chirp 3813 ms
leaders plus Google models · 15 other models in the full table
Word Error Rate% · lower is better · Google models markedEvery measured STT model on Word Error Rate, with Google's models highlighted.
  1. #1Universal 3.5 Pro3.2%
  2. #2resonant-13.4%
  3. #3Chirp 34.1%
  4. #4Enhanced4.3%
  5. #5GPT-4o Transcribe4.7%
  6. #6Grok STT4.8%
  7. #7STT 14.8%
  8. #9Chirp 25.0%
leaders plus Google models · 20 other models in the full table
Time to First Tokenms · lower is better · Google models markedEvery measured STT model on Time to First Token, with Google's models highlighted.
  1. #1Universal 3.5 Pro1030 ms
  2. #2Flux1089 ms
  3. #3Flux Multilingual1175 ms
  4. #4Parakeet TDT 0.6B v31210 ms
  5. #5Whisper Large v31241 ms
  6. #6Nova 31434 ms
  7. #7Nova 21436 ms
  8. #23Chirp 35999 ms
  9. #24Chirp 26000 ms
leaders plus Google models · 15 other models in the full table
TTFS vs WER — Google vs every measured modelEach point is one measured STT model · Google models highlighted · 30-day averagesTime to Final Segment against Word Error Rate for every measured STT model, with Google's models highlighted.
STT RT v5 — 64 ms, 5.8%Parakeet TDT 0.6B v3 — 70 ms, 11.0%STT 1 — 83 ms, 4.8%Nova 3 — 99 ms, 6.3%Nova 2 — 101 ms, 7.9%Ink 2 — 108 ms, 4.9%Scribe v2 Realtime — 120 ms, 5.5%Universal 3.5 Pro — 146 ms, 3.2%Default via Azure — 155 ms, 5.3%Whisper Large v3 — 180 ms, 8.2%Velma 2 STT Streaming — 191 ms, 6.4%Grok STT — 199 ms, 4.8%Pulse — 205 ms, 5.3%Default via Speechmatics — 223 ms, 5.5%Default via Gradium — 249 ms, 10.0%resonant-1 — 288 ms, 3.4%Enhanced — 341 ms, 4.3%Voxtral Mini Transcribe Realtime 2602 — 356 ms, 6.2%GPT Realtime Whisper — 559 ms, 5.2%GPT-4o mini Transcribe — 623 ms, 5.1%Solaria 1 — 680 ms, 7.1%GPT-4o Transcribe — 742 ms, 4.7%Chirp 2 — 811 ms, 5.0%Chirp 3 — 813 ms, 4.1%
Benchmarked models with their Time to Final Segment over the last 30 days.
ModelHostTTFSRank
Chirp 2Google811 ms23rd of 24
Chirp 3Google813 ms24th of 24

Text-to-Speech

Full TTS dashboard

Ranked on Time to First Audio against 30 measured models.

Time to First Audioms · lower is better · Google models markedEvery measured TTS model on Time to First Audio, with Google's models highlighted.
  1. #1Palabra TTS v1116 ms
  2. #2vui124 ms
  3. #3TTS Flash 2128 ms
  4. #4TTS 2176 ms
  5. #5Blizzard235 ms
  6. #6Neural236 ms
  7. #7Mist v3256 ms
  8. #24Chirp 3 HD512 ms
leaders plus Google models · 22 other models in the full table
Word Error Rate% · lower is better · Google models markedEvery measured TTS model on Word Error Rate, with Google's models highlighted.
  1. #1TTS RT v13.7%
  2. #2TTS Rt v23.9%
  3. #3Neural4.3%
  4. #4Eleven v3 Conversational4.4%
  5. #5Lightning v3.1 Pro4.4%
  6. #6Default4.6%
  7. #7S2.1 Pro4.6%
  8. #18Chirp 3 HD5.2%
leaders plus Google models · 22 other models in the full table
TTFA vs WER — Google vs every measured modelEach point is one measured TTS model · Google models highlighted · 30-day averagesTime to First Audio against Word Error Rate for every measured TTS model, with Google's models highlighted.
Palabra TTS v1 — 116 ms, 5.9%vui — 124 ms, 7.9%TTS Flash 2 — 128 ms, 4.9%TTS 2 — 176 ms, 4.8%Blizzard — 235 ms, 7.4%Neural — 236 ms, 4.3%Mist v3 — 256 ms, 6.3%TTS Rt v2 — 262 ms, 3.9%Sonic 3.5 — 274 ms, 6.1%TTS RT v1 — 275 ms, 3.7%Dragon HD Latest — 310 ms, 5.1%Coda — 313 ms, 5.2%Aura 2 — 328 ms, 5.3%S2.1 Pro — 374 ms, 4.6%Default — 384 ms, 4.6%Speech 2.8 Turbo — 411 ms, 4.8%Eleven v3 Conversational — 412 ms, 4.4%Grok TTS — 420 ms, 4.7%S1 — 434 ms, 4.9%Flash v2.5 — 455 ms, 6.7%Speech 2.8 HD — 460 ms, 4.7%Sonic 3.6 — 466 ms, 5.5%Simba 3.2 — 484 ms, 4.7%Simba 3.0 — 526 ms, 5.3%Falcon 2 — 549 ms, 5.2%Lightning v3.1 Pro — 586 ms, 4.4%Qwen3 TTS Flash Realtime — 645 ms, 8.8%S2.1 Pro Free — 806 ms, 4.7%GPT-4o mini TTS — 1075 ms, 4.8%Chirp 3 HD — 512 ms, 5.2%
Benchmarked models with their Time to First Audio over the last 30 days.
ModelHostTTFARank
Chirp 3 HDGoogle512 ms24th of 30

Speech-to-Speech

Full S2S dashboard

Ranked on Voice-to-Voice Latency against 2 measured models.

Voice-to-Voice Latencyms · lower is better · Google models markedEvery measured S2S model on Voice-to-Voice Latency, with Google's models highlighted.
  1. #1GPT Realtime 21306 ms
  2. #2Gemini 3.1 Flash Live (Preview)1380 ms
Instruction Adherence% · higher is better · Google models markedEvery measured S2S model on Instruction Adherence, with Google's models highlighted.
  1. #1Gemini 3.1 Flash Live (Preview)76%
  2. #2GPT Realtime 271%
V2V vs Instruction — Google vs every measured modelEach point is one measured S2S model · Google models highlighted · 30-day averagesVoice-to-Voice Latency against Instruction Adherence for every measured S2S model, with Google's models highlighted.
GPT Realtime 2 — 1306 ms, 71%Gemini 3.1 Flash Live (Preview) — 1380 ms, 76%
Benchmarked models with their Voice-to-Voice Latency over the last 30 days.
ModelHostV2VRank
Gemini 3.1 Flash Live (Preview)Google1380 ms2nd of 2

Limits of this comparison

  • The measured endpoints do not represent every Google Cloud region, Chirp voice, Gemini capability or language.

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