Leaderboard
Results by accent
Every number on this page comes from real, reproducible runs. Anything we have not measured yet is marked “Coming soon”.
“¿Quieres que te mande el artículo?”
- 1open source
whisper-small
faster-whisper · 20 samples
WER
7.5%
p50
1.05 s
CER
3.3%
- 2open source
whisper-large-v3-turbo
faster-whisper · 20 samples
WER
22.6%
p50
3.82 s
CER
15.5%
- 3open source
whisper-base
faster-whisper · 20 samples
WER
26.3%
p50
480 ms
CER
16.5%
- 4open source
whisper-tiny
faster-whisper · 20 samples
WER
27.4%
p50
372 ms
CER
10.7%
| # | Model | Error (WER) ↓ | Latency p50 ↓ | CER | Price |
|---|---|---|---|---|---|
| 1 | whisper-small faster-whisper · int8 · CPU · 20 samples | 7.5% | 1.05 s | 3.3% | Free · open source |
| 2 | whisper-large-v3-turbo faster-whisper · int8 · CPU · 20 samples | 22.6% | 3.82 s | 15.5% | Free · open source |
| 3 | whisper-base faster-whisper · int8 · CPU · 20 samples | 26.3% | 480 ms | 16.5% | Free · open source |
| 4 | whisper-tiny faster-whisper · int8 · CPU · 20 samples | 27.4% | 372 ms | 10.7% | Free · open source |
Coming soon · STT
AssemblyAI Universal
From US$0.15 / hour
Deepgram Nova-3
US$0.0077 / min (streaming)
Speechmatics
From ~US$0.24 / hour
Gladia
US$0.61 / hour (async)
Google Cloud STT (Chirp)
US$0.016 / min (streaming)
OpenAI gpt-4o-mini-transcribe
US$0.003 / min
Commercial providers: we will measure them once we can do so with our own accounts and within their terms. Until then we show only their official price, never estimated metrics. Latency = total time to transcribe one sentence (CPU).
- 1open source
es_AR-daniela-high
Piper · 20 samples
WER
9.7%
p50
352 ms
CER
4.8%
¿Quieres que te mande el artículo?
| # | Model | Error (WER) ↓ | Latency p50 ↓ | CER | Price |
|---|---|---|---|---|---|
| 1 | es_AR-daniela-high Piper · 20 samples ¿Quieres que te mande el artículo? | 9.7% | 352 ms | 4.8% | Free · open source |
Coming soon · TTS
ElevenLabs
Paid plan required for commercial use
Google Cloud TTS (Chirp 3 HD)
US$30 / 1M characters
Azure AI Speech (Neural)
US$16 / 1M characters
Amazon Polly (Neural)
US$16 / 1M characters
OpenAI tts-1
US$15 / 1M characters
Deepgram Aura-2
US$30 / 1M characters
Cartesia Sonic
See official site
Commercial providers: we will measure them once we can do so with our own accounts and within their terms. Until then we show only their official price, never estimated metrics. TTS scored for intelligibility: audio is transcribed with whisper-large-v3-turbo and compared with the text. Latency = time until a sentence's full audio is ready (CPU).