Start building →OverviewMultilingualCode-switchingEmotion

Speech-to-Text · Multilingual

Indonesian

Language

Request a new language

Tell us which language you want added to this benchmark.

Benchmarks

X
Y
20ms50ms100ms200ms500ms1.00s2.00s5.0%10.0%15.0%Finalize latency lower is better · log scaleWER lower is betterGoogle Gemini 3.5 Transcribe LivexAI Grok STTSoniox stt-rt-v5ElevenLabs Scribe v2 RealtimeAlibaba Qwen3-ASR RealtimeNari Qwen3-ASR FastOpenAI GPT Live TranscribeNari Qwen3-ASROpenAI GPT Realtime WhisperGladia Solaria-1Deepgram Nova-3speko.ai/benchmarks

WER

Share of words the transcript got wrong.

GoogleGemini 3.5 Transcribe LiveTry →2.6%
MetaMetaMuse Voice TranscribeTry →3.6%
Sonioxstt-rt-v5Try →3.6%
ElevenLabsScribe v2 RealtimeTry →4.2%
AlibabaQwen3-ASR RealtimeTry →4.4%
xAIGrok STTTry →4.6%
OpenAIGPT Live TranscribeTry →5.7%
NariNariQwen3-ASR FastTry →5.7%
OpenAIGPT Realtime Whisper5.8%
NariNariQwen3-ASRTry →5.8%
ModulateModulateVelma 2 multilingualTry →7.1%
CartesiaInk-WhisperTry →9.2%
GladiaSolaria-1Try →9.9%
DeepgramNova-3Try →14.6%
0.0%5.0%10.0%15.0%20.0%

Finalize latency

lower is better

Time from the caller stopping to the finalized transcript, not the first partial — the agent cannot reply until it lands. Measured on English audio; Indonesian has no sweep of its own yet.

NariNariQwen3-ASRTry →25ms · 43ms
NariNariQwen3-ASR FastTry →33ms · 45ms
DeepgramNova-3Try →189ms · 770ms
Sonioxstt-rt-v5Try →340ms · 1.33s
OpenAIGPT Live TranscribeTry →620ms · 770ms
OpenAIGPT TranscribeTry →817ms · 1.17s
OpenAIGPT-4o mini TranscribeTry →822ms · 1.09s
OpenAIGPT-4o TranscribeTry →848ms · 1.18s
AlibabaQwen3-ASR RealtimeTry →889ms · 1.21s
GladiaSolaria-1Try →910ms · 3.10s
xAIGrok STTTry →1.13s · 1.34s
OpenAIGPT Realtime Whisper1.14s · 2.26s
GoogleGemini 3.5 Transcribe LiveTry →1.28s · 1.54s
ElevenLabsScribe v2 RealtimeTry →1.75s · 1.98s
0ms300ms600ms900ms1.20s

p50 p90

Time to first token

lower is better

How long before the first word of the transcript arrives. Measured on English audio; Indonesian has no sweep of its own yet.

DeepgramNova-3Try →976ms · 1.98s
xAIGrok STTTry →1.03s · 2.01s
Sonioxstt-rt-v5Try →1.53s · 2.04s
GladiaSolaria-1Try →1.61s · 3.65s
MetaMetaMuse Voice TranscribeTry →1.62s · 2.23s
OpenAIGPT Live TranscribeTry →1.72s · 2.24s
NariNariQwen3-ASR FastTry →1.74s · 1.74s
NariNariQwen3-ASRTry →1.75s · 1.78s
GoogleGemini 3.5 Transcribe LiveTry →1.80s · 2.30s
OpenAIGPT Realtime Whisper1.81s · 2.30s
ElevenLabsScribe v2 RealtimeTry →2.22s · 2.36s
OpenAIGPT TranscribeTry →6.28s · 10.9s
OpenAIGPT-4o mini TranscribeTry →6.33s · 10.6s
OpenAIGPT-4o TranscribeTry →6.40s · 10.8s
AlibabaQwen3-ASR RealtimeTry →6.81s · 7.50s
0ms2.00s4.00s6.00s8.00s

p50 p90

Cost

lower is better

$ per 1,000 minutes of audio · vendor list price, the same in every language

$1.00
$2.00
$2.00
$2.10
$3.00
$3.70
$4.50
$4.80
$5.40
$6.00
$6.50
$9.00
$12.50
$16.00
$17.00
$17.00
NariNariQwen3-ASRNariNariQwen3-ASR FastSonioxstt-rt-v5AlibabaQwen3-ASRMetaMetaMuse Voice TranscribeElevenLabsScribe v2OpenAIGPT TranscribeDeepgramNova-3AlibabaQwen3-ASR RealtimeOpenAIGPT-4o TranscribeElevenLabsScribe v2 RealtimeGoogleGemini 3.5 Transcribe LiveGladiaSolaria-1GoogleChirp 3OpenAIGPT Live TranscribeOpenAIGPT Realtime Whisper