Portuguese speech to text, built on PyAI Hear
Portuguese speech to text turns spoken Portuguese into written text your software can store, search and act on. PyAI Hear is the recognition engine underneath, the same one running live phone agents on PyAI. Hear reads clear studio audio and hard 8 kHz phone lines, follows Portuguese across Brazil and Portugal, and returns text as plain, JSON or timestamped output. You call it from your own code and the audio stays in your own flow. Drop in a recording and PyAI works as a Portuguese voice to text tool, or call the same endpoint to transcribe Portuguese audio to text from your own code. The same audio to text generator serves a single upload or thousands of calls a day. PyAI works as a free Portuguese speech to text generator, so a saved file or a quick recording turns into text you can edit, search and reuse.
Key takeaways
- Price: $0.001/min, billed by the whole minute.
- Varieties: Brazilian Portuguese and European Portuguese from Portugal run on one model, so a line that mixes callers from São Paulo and Lisbon still returns a single clean transcript with no variety selected in advance.
- No lock in: no platform fee.
- Accuracy: word error rate is 1.59% on clean audio and 4.76% on real 8 kHz phone audio, PyAI's own benchmark numbers measured on English and reported as industry parity rather than a lead.
- Speed: first partial near 200 ms, long files at 92 to 247 times real time.
- Drop in: a whisper-1 client reaches Hear by swapping the base URL, with single files up to 25 MB per call.
How Portuguese transcription works
You send Portuguese audio to one endpoint and get text back. The same engine runs as a file call, a live stream, or a background job.
Post a Portuguese file to the transcription endpoint, or open a WebSocket and stream PCM16 frames for a live support line or meeting.
The engine runs in one region, marks speaker turns, and commits each final line once a speaker finishes, so partial text firms up as more audio arrives.
Smart formatting, dictation and filler removal clean the text, and numbers, dates and currency are written the way a Portuguese reader expects.
Plain text, JSON, or verbose JSON with per segment timestamps and speaker turns, ready for notes, captions or search.
What Hear handles in Portuguese
Portuguese brings its own accents, numbers and names. Here is how Hear deals with them.
Brazil and Portugal
Brazilian Portuguese with its open vowels and European Portuguese with its reduced, clipped speech both run on one model. You do not pick a variety first, and a call mixing the two still resolves into one transcript.
Phone audio
Hear is tuned for 8 kHz phone lines, the hard case. PyAI measures 4.76% word error rate on noisy, accented 8 kHz English audio, so test your own Portuguese calls before launch.
Nasal and reduced sounds
Portuguese nasal vowels and the reduced unstressed syllables common in European speech are handled, so fast or softly spoken words still come back readable.
Numbers and names
Amounts, dates and order numbers come out in digits, and up to 5 custom vocabulary terms teach it the names, places and product words your business uses most.
Overlap and turns
A turn detector tags who spoke when, so two voices over each other return as separate labeled segments rather than one run on block.
Long recordings
A long recording is split into overlapping chunks, transcribed in parallel and stitched back, which is how throughput reaches 92 to 247 times real time.
Accuracy and speed
Hear is at industry parity for accuracy, held on the hard case of real phone audio. The figures below are PyAI's own measurements on English speech, not a per language claim.
What you can transcribe in Portuguese
The same Portuguese endpoint feeds notes, captions, search and analytics.
Call transcription
Log every Portuguese support or sales call as searchable text, then send it to quality review, coaching or a CRM note.
Order and delivery support
Capture tracking and delivery calls as text, so order numbers, dates and promises are written down instead of lost after the call ends.
Meeting notes
Stream a Portuguese meeting live and get a running transcript with speaker labels, so action items are caught without a note taker.
Captions and subtitles
Generate Portuguese captions for video with per segment timestamps that drop straight into an SRT or VTT file.
Interviews and podcasts
Transcribe long Portuguese audio as a background job and get timestamped text back to quote, clip or index.
Voice search
Send short Portuguese clips and act on the returned text, so a user speaks a query or command inside your app.
Formats, limits and output
| Input | Audio file up to 25 MB for a single file call, or a live PCM16 stream over WebSocket for real time. OpenAI compatible, so a whisper-1 client works unchanged. |
|---|---|
| Output | Plain text, JSON, or verbose JSON with per segment timestamps. Smart formatting, dictation and filler removal are applied on the way out. |
| Modes | Sync file call for short clips, live streaming for calls and meetings, and background jobs for long recordings that run 92 to 247 times real time. |
| Tuning | Set the language to Portuguese, add numerals handling, and load up to 5 custom vocabulary terms so names and product words land right. |
| Billing | $0.001/min of audio, rounded to the whole minute. An empty transcript costs nothing, and a busy engine returns a retry, not a charge. |
Call Portuguese speech to text from your own code
Hear follows the OpenAI transcription shape, so most Portuguese code reaches PyAI by swapping the base URL. Full docs live at PyAI SDKs.
SDK
From a server, the SDK posts a Portuguese file or opens a stream and returns the transcript in a few lines. The SDK speaks the whisper-1 shape, so transcription code you already run moves across by swapping the base URL.
MCP
With the MCP server, an agent can transcribe a Portuguese recording on its own inside a flow, for example turning a support call into a ticket note, without you wiring the API by hand.
CLI
From the terminal, the CLI takes a Portuguese audio file or a folder and writes out text or timestamped JSON, handy for clearing a backlog of recordings in one pass.
Where Hear fits
PyAI is voice infrastructure. Hear, Speak and Omni run together in one stack and one region, so recognition, reasoning and reply keep a single session instead of hopping between vendors. Hear is not a consumer dictation app, and PyAI does not resell another company's recognizer. Hear is the Portuguese transcription engine you call from your own product, on your own calls and files. See the published numbers on PyAI benchmarks.
Portuguese speech to text FAQ
How accurate is Portuguese speech to text on Hear?
Hear measures 1.59% word error rate on clean audio and 4.76% on real 8 kHz phone audio. Both are PyAI's own benchmark figures, taken on English and at industry parity, so they describe the engine overall and not any single Portuguese variety or recording quality.
Does it handle both Brazilian and European Portuguese?
Yes. Brazilian Portuguese and European Portuguese from Portugal both run on one model, so you do not choose a variety in advance. Up to 5 custom vocabulary terms help with names and product words that are specific to your audio.
How much does it cost to transcribe Portuguese audio?
Hear charges $0.001/min of audio, rounded to the whole minute, and an empty transcript is free. A new account opens with $50 of credit and no credit card, enough for plenty of test audio before you spend anything, and sandbox keys never bill.
Can I reuse my whisper-1 code?
Yes. Hear follows the OpenAI transcription format, so a whisper-1 client works once it points at the PyAI base URL. You keep the same request shape and get text, JSON or verbose JSON with timestamps back as before, so migration takes minutes of work, not a rewrite.
Does Hear handle live Portuguese audio?
Both live and files. A single call takes clips up to 25 MB, a WebSocket stream handles live calls and meetings with partials near 200 ms, and a background job runs long recordings at 92 to 247 times real time.
Prices checked November 2026. Hear figures from pyai.com/benchmarks.