Catalog · 41 models · 14 providers
Every speech-to-text model worth using, one rate card.
1
Total models
20
Languages
$0.003
Cheapest /min
18.2%
Best WER
1 models
filtered from 41
Rev AI
Rev AI Reverb — async speech-to-text built on Rev's Reverb ASR model (trained on 3M+ hours of human-transcribed audio), with speaker diarization for up to 8 speakers and word-level timestamps. English-core with broad language support via the async API.
WER
CER
Speed Factor
0.2x
Uptime
Billing
per 1s
Retention
20 languages
(view all)
Every pairing we index, grouped by model. Same audio, same scoring.
Google Cloud
$0.0107/min · 9.95% WER
Deepgram
$0.0077/min · 20.98% WER
Gladia
$0.0101/min · 9.80% WER
Amazon Web Services
$0.006/min · 13.02% WER
Deepgram
$0.0058/min · 16.54% WER
ElevenLabs
$0.004/min · 9.15% WER
Alibaba (Qwen)
$0.0021/min · 12.27% WER
Mistral (Voxtral)
$0.003/min · 12.70% WER
Groq
$0.0007/min · 14.31% WER
Groq
$0.0019/min · 14.41% WER
OpenAI
$0.006/min · 17.67% WER
OpenAI
$0.003/min · 13.58% WER
OpenAI
$0.006/min · 14.61% WER
Soniox
$0.0017/min · 8.73% WER
Accuracy, latency and pricing, checked against what providers actually ship.
What Amazon Transcribe Medical offers in 2026: features, specs, pricing vs Google and Nuance, HIPAA posture, research clues, and where it falls short.
July 3, 2026
Read
What Google Cloud Chirp 3 actually is: release timeline, WER and Elo benchmarks, pricing, specs, known issues, and how it compares to Azure and ElevenLabs.
July 3, 2026
Read
Where Deepgram Base fits in 2026: API behavior, variants, latency, specs, limitations, and when to pick Nova-3 or Flux instead.
July 3, 2026
Read
