Voice AI BenchmarksBy

PROVIDERLAST 30 DAYS

official resource

Alibaba voice AI models and benchmarks

Alibaba lists 5 STT and TTS models in Coval. Fastest dated mean latency over 30 days: STT: Qwen3 ASR 1.7b on Baseten (dedicated inference) at 39 ms TTFS, with 2.9% WER. TTS: Qwen3 TTS Fast on Nari Labs at 68 ms TTFA, with 1.8% WER. Last measured .

Alibaba Cloud provides Model Studio APIs for Qwen-family generative models, including real-time speech synthesis, and releases open-weight Qwen3 speech models.

Measured models
5
STTTTS
Best STT model#1 / 29
39ms
Qwen3 ASR 1.7b
Best TTS model#3 / 32
68ms
Qwen3 TTS Fast

Overview

Alibaba Cloud serves its TTS first-party from an Asia-region deployment, which makes serving geography an important part of latency interpretation.

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
Qwen3 ASR 1.7b, Qwen3 ASR Fast
Text-to-Speech
Qwen3 TTS Flash Realtime, Qwen3 TTS 1.7b, Qwen3 TTS Fast

Speech-to-Text

Full STT dashboard

Ranked on Time to Final Segment against 29 measured models.

Time to Final Segmentms · lower is better · Alibaba models markedEvery measured STT model on Time to Final Segment, with Alibaba's models highlighted.
  1. #1Qwen3 ASR 1.7b39 ms
  2. #2STT RT v545 ms
  3. #3Qwen3 ASR Fast47 ms
  4. #4STT 166 ms
  5. #5Nova 385 ms
  6. #6Nova 287 ms
  7. #7Parakeet TDT 0.6B v397 ms
leaders plus Alibaba models · 22 other models in the full table
Word Error Rate% · lower is better · Alibaba models markedEvery measured STT model on Word Error Rate, with Alibaba'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%
leaders plus Alibaba models · 24 other models in the full table
Time to First Tokenms · lower is better · Alibaba models markedEvery measured STT model on Time to First Token, with Alibaba'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 11211 ms
  8. #17Qwen3 ASR Fast1752 ms
leaders plus Alibaba models · 19 other models in the full table
TTFS vs WER — Alibaba vs every measured modelEach point is one measured STT model · Alibaba models highlighted · 30-day averagesTime to Final Segment against Word Error Rate for every measured STT model, with Alibaba's models highlighted.
STT RT v5 — 45 ms, 4.8%STT 1 — 66 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 — 148 ms, 3.6%Universal 3.5 Pro — 168 ms, 2.4%Whisper Large v3 via Together AI — 200 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 — 410 ms, 5.4%GPT Realtime Whisper — 599 ms, 4.0%Solaria 1 — 636 ms, 6.0%GPT-4o mini Transcribe — 708 ms, 3.9%Chirp 3 — 768 ms, 3.2%GPT-4o Transcribe — 783 ms, 3.8%Chirp 2 — 908 ms, 3.9%Qwen3 ASR 1.7b — 39 ms, 2.9%Qwen3 ASR Fast — 47 ms, 2.6%
Benchmarked models with their Time to Final Segment over the last 30 days.
ModelHostTTFSRank
Qwen3 ASR 1.7bBaseten39 ms1st of 29
Qwen3 ASR FastNari Labs47 ms3rd 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 · Alibaba models markedEvery measured TTS model on Time to First Audio, with Alibaba's models highlighted.
  1. #1vui60 ms
  2. #2Default63 ms
  3. #3Qwen3 TTS Fast68 ms
  4. #4TTS Flash 292 ms
  5. #5Qwen3 TTS 1.7b107 ms
  6. #6Simba 3.0111 ms
  7. #7Palabra TTS v1119 ms
  8. #30Qwen3 TTS Flash Realtime718 ms
leaders plus Alibaba models · 24 other models in the full table
Word Error Rate% · lower is better · Alibaba models markedEvery measured TTS model on Word Error Rate, with Alibaba's models highlighted.
  1. #1Sonic 3.61.5%
  2. #2S2.1 Pro Free1.5%
  3. #3Simba 3.01.5%
  4. #4Chirp 3 HD1.5%
  5. #5S2.1 Pro1.5%
  6. #6Lightning v3.1 Pro1.5%
  7. #7Sonic 3.51.5%
  8. #16Qwen3 TTS 1.7b1.8%
  9. #19Qwen3 TTS Fast1.8%
  10. #31Qwen3 TTS Flash Realtime3.0%
leaders plus Alibaba models · 22 other models in the full table
TTFA vs WER — Alibaba vs every measured modelEach point is one measured TTS model · Alibaba models highlighted · 30-day averagesTime to First Audio against Word Error Rate for every measured TTS model, with Alibaba's models highlighted.
vui — 60 ms, 2.1%Default — 63 ms, 1.9%TTS Flash 2 — 92 ms, 1.7%Simba 3.0 — 111 ms, 1.5%Palabra TTS v1 — 119 ms, 2.6%Simba 3.2 — 126 ms, 1.9%drama3-preview — 161 ms, 2.1%Flash v2.5 — 191 ms, 1.8%TTS 2 — 192 ms, 1.6%Eleven v4 Turbo — 201 ms, 1.7%Flux TTS — 205 ms, 1.7%TTS RT v1 — 254 ms, 2.4%TTS Rt v2 — 256 ms, 2.4%Mist v3 — 287 ms, 2.1%Sonic 3.5 — 293 ms, 1.5%Phantom Z 3.4 conversational — 295 ms, 1.8%Aura 2 — 297 ms, 1.8%Sonic 3.6 — 299 ms, 1.5%Falcon 2 — 303 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 — 392 ms, 1.5%Daytona TTS — 486 ms, 1.6%Aura 2 En — 572 ms, 2.1%Chirp 3 HD — 598 ms, 1.5%S2.1 Pro Free — 850 ms, 1.5%GPT-4o mini TTS — 2583 ms, 1.8%Qwen3 TTS Fast — 68 ms, 1.8%Qwen3 TTS 1.7b — 107 ms, 1.8%Qwen3 TTS Flash Realtime — 718 ms, 3.0%
Benchmarked models with their Time to First Audio over the last 30 days.
ModelHostTTFARank
Qwen3 TTS Flash RealtimeAlibaba718 ms30th of 32
Qwen3 TTS 1.7bBaseten107 ms5th of 32
Qwen3 TTS FastNari Labs68 ms3rd of 32

How fast are Alibaba's STT and TTS models?

Qwen3 ASR 1.7b on Baseten (dedicated inference) measures mean 39 ms time to final segment (1st of 29) among STT systems. Last measured 2026-10-14. Qwen3 ASR Fast on Nari Labs measures mean 47 ms time to final segment (3rd of 29) among STT systems. Last measured 2026-10-14. Qwen3 TTS Fast on Nari Labs measures mean 68 ms time to first audio (3rd of 32) among TTS systems. Last measured 2026-10-14. Qwen3 TTS 1.7b on Baseten (dedicated inference) measures mean 107 ms time to first audio (5th of 32) among TTS systems. Last measured 2026-10-14. Qwen3 TTS Flash Realtime measures mean 718 ms time to first audio (30th of 32) among TTS systems. Last measured 2026-10-14.

How accurate are Alibaba's STT and TTS models?

Qwen3 ASR Fast on Nari Labs measures 2.6% word error rate (3rd of 31) among STT systems. Last measured 2026-10-14. Qwen3 ASR 1.7b on Baseten (dedicated inference) measures 2.9% word error rate (5th of 31) among STT systems. Last measured 2026-10-14. Qwen3 TTS 1.7b on Baseten (dedicated inference) measures 1.8% word error rate (16th of 32) among TTS systems. Last measured 2026-10-14. Qwen3 TTS Fast on Nari Labs measures 1.8% word error rate (19th of 32) among TTS systems. Last measured 2026-10-14. Qwen3 TTS Flash Realtime measures 3.0% word error rate (31st of 32) among TTS systems. Last measured 2026-10-14.

Which Alibaba model is fastest?

Its fastest dated STT result is Qwen3 ASR 1.7b on Baseten (dedicated inference) at mean 39 ms time to final segment (1st of 29) among STT systems, with 2.9% WER. Last measured 2026-10-14. Its fastest dated TTS result is Qwen3 TTS Fast on Nari Labs at mean 68 ms time to first audio (3rd of 32) among TTS systems, with 1.8% WER. Last measured 2026-10-14.

Limits of this comparison

Coval measures the Qwen3 TTS Flash endpoint served by Alibaba Cloud, plus the open-weight Qwen3-ASR 1.7B and Qwen3-TTS 1.7B models served by Baseten.

  • Coval runs the benchmark from us-east-1 rather than an Asia-local worker. The results cover the measured Qwen3 endpoints, not Alibaba Cloud's wider model platform or every Qwen audio release.

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