Transcribe audio for $0.006 per minute — native Portuguese (PT-BR), word-level timestamps, speaker diarization, and pay-as-you-go billing with no monthly minimum.
Brainiall’s speech-to-text API costs $0.006 per minute of audio — the same list price as the OpenAI Whisper API and 2.7–4x below Google, AWS, and Azure — with pay-as-you-go billing, no monthly minimum, word-level timestamps included, and Portuguese (PT-BR) as a first-class language.
$0.006 per minute, billed by usage. No subscription, no monthly minimum, $10 in free credits (about 27 hours of audio) to start.
| Brainiall | OpenAI Whisper API | Deepgram | AssemblyAI | |
|---|---|---|---|---|
| Price per minute (batch) | $0.006 | $0.006 | from $0.0043 (Nova-3) | from $0.0061 (Universal) |
| Word-level timestamps | Included | Limited (verbose output) | Included | Included |
| Speaker diarization | Dedicated endpoint | N/A | Paid add-on | Paid add-on |
| Portuguese (PT-BR) | First-class language | Supported | Supported | Supported |
| Monthly minimum | None | None | None (prepaid credits) | None |
| Free credits to start | $10 | N/A | Trial credits | Trial credits |
Competitor prices are list prices published on each provider’s public pricing page, checked July 2026. Brainiall prices are the live pay-as-you-go rates. Google Speech-to-Text ($0.016/min), AWS Transcribe ($0.024/min), and Azure Speech ($0.016/min) sit above every provider in this table.
curl -X POST https://api.brainiall.com/v1/stt/transcribe \
-H "Authorization: Bearer $BRAINIALL_API_KEY" \
-F "audio=@meeting.mp3" \
-F "include_timestamps=true"Send any WAV, MP3, OGG, FLAC, or WebM file as multipart form data. The response includes the transcript plus per-word start/end times and confidence scores. Get a key at app.brainiall.com — first call in about 60 seconds.
Managed speech-to-text list prices in 2026 range from about $0.0043 per minute (Deepgram Nova-3 batch) to $0.024 per minute (AWS Transcribe). Brainiall and the OpenAI Whisper API both charge $0.006 per minute. Brainiall adds word-level timestamps, a diarization endpoint, Portuguese as a first-class language, and pay-as-you-go billing with no monthly minimum — so the effective cost for a production pipeline is among the lowest available.
1,000 hours is 60,000 minutes. At Brainiall’s $0.006 per minute that is $360. The same volume costs about $960 on Google Speech-to-Text ($0.016/min) and about $1,440 on AWS Transcribe ($0.024/min) at list prices.
Yes. Portuguese is a first-class language: models are selected and evaluated against Brainiall’s own Portuguese benchmark suite, not treated as a long-tail locale. English, Spanish, and other major languages are supported as well.
Speaker diarization (who spoke when) is available as a dedicated endpoint, /v1/audio/diarize. You can combine its speaker turns with the word-level timestamps from /v1/stt/transcribe to produce speaker-labeled transcripts.
Every new account gets $10 in free credits — about 1,600 minutes (roughly 27 hours) of transcription — with no credit card required.
Create an account at app.brainiall.com, copy your key from the dashboard, and call https://api.brainiall.com/v1/stt/transcribe with an Authorization: Bearer header. Most developers make their first successful call in about a minute.
Get a free API key with $10 in credits — about 27 hours of audio. No credit card, no subscription, cancel anytime.