Brazilian and European Portuguese differ so much in phonology that they are harder to cross-transcribe than Spanish and Italian. Our AI handles the extreme vowel reduction of Lisbon speech, Brazilian open vowels, clitic pronoun positions, and the inconsistently adopted 1990 Spelling Agreement.
Drop your file here or click to browse
.mp3, .wav, .m4a, .aac, .ogg, .flac, .mp4, .mov, .avi, .mkv, .webm·up to 500MB
Brazilian and European Portuguese have diverged so far phonologically that a tool trained primarily on one variety will produce poor results on the other. European Portuguese reduces unstressed vowels so aggressively that the word despertar sounds like /dshprtar/ to Brazilian ears. Brazilian Portuguese preserves open vowels and adds epenthetic vowels between consonant clusters, so ritmo becomes /hitchimu/. Beyond pronunciation, the two varieties split on clitic pronoun placement (me diz in Brazil vs diz-me in Portugal), contrastive nasal vowels that change meaning, and the 1990 Spelling Agreement that Portugal adopted but Brazil implemented differently. Vocova’s AI detects the variety automatically and applies the correct phonological, grammatical, and orthographic rules.
Drag and drop any recording containing Portuguese speech. The AI detects whether the speaker uses Brazilian or European Portuguese from the first few seconds of audio.
The engine adapts to the detected variety — recovering heavily reduced European vowels, handling Brazilian open vowel patterns, placing clitic pronouns correctly, and applying the appropriate spelling conventions.
Review the transcript with all nasal vowel markers, accent marks, and variety-specific spelling in place. Export in your preferred format with timestamps and speaker labels.
The AI identifies the Portuguese variety from acoustic features within seconds — Brazilian open vowels and palatalized consonants vs European closed vowels and sibilant s-sounds — and applies the correct transcription rules for that variety throughout the recording.
European Portuguese reduces unstressed vowels so aggressively that entire syllables seem to disappear. The word telefonema sounds like /tlfnema/, and despertar like /dshprtar/. The AI recovers the full written form from these heavily reduced acoustic signals, producing correct spelling despite minimal vowel information.
Portuguese nasal vowels change meaning: mão (hand) vs mau (bad), lã (wool) vs lá (there), sim (yes) vs si (musical note). The AI distinguishes nasal from oral vowels accurately and places the tilde (ã, õ) correctly in the written output.
Brazilians place clitic pronouns before the verb (me diz, me conta) while Portuguese speakers place them after (diz-me, conta-me). This is not a minor stylistic preference — it is a fundamental grammatical difference. The AI detects the variety and applies the correct clitic position.
The 1990 Orthographic Agreement changed spellings like acção to ação and óptimo to ótimo, but adoption has been inconsistent between Brazil and Portugal. The AI applies the currently standard spelling for the detected variety rather than forcing a single standard across both.
Get transcripts that sound right for the variety spoken. Brazilian recordings produce Brazilian spellings and clitic positions; European recordings produce European conventions. No cross-contamination between varieties.
European Portuguese vowel reduction makes it one of the most phonologically compressed Romance languages. The AI produces fully spelled, readable text from speech that sounds radically different from its written form.
Nasal vowels are contrastive in Portuguese and their written markers (ã, õ, the m/n before consonants) must be placed correctly. The AI handles these distinctions accurately in both varieties.
When a meeting includes Brazilian and Portuguese speakers, each voice is labeled separately. Vocabulary and grammatical differences between varieties (autocarro vs ônibus, pequeno-almoço vs café da manhã) are transcribed as spoken.
Transcribe Globo broadcasts, Brazilian YouTube content, and podcasts with correct Brazilian Portuguese conventions — open vowels, proclitic pronouns, and Brazilian vocabulary accurately rendered.
Get accurate transcripts from RTP, Portuguese radio, and interviews despite the extreme vowel reduction that makes European Portuguese challenging for speech recognition.
Document meetings and calls across Brazil, Portugal, Angola, and Mozambique. The AI adapts to each speaker’s variety and produces transcripts with the correct regional conventions.
See how spoken Portuguese maps to its written form in both varieties. Especially valuable for European Portuguese, where vowel reduction makes the connection between sound and spelling difficult to perceive.
The AI identifies the variety from acoustic features in the first seconds of speech: Brazilian Portuguese has open unstressed vowels and palatalized /t/ and /d/ before /i/, while European Portuguese has heavily reduced vowels and alveolar sibilants. This detection is automatic — no manual dialect selection is needed.
The AI applies the currently standard spelling for the detected variety. For Brazilian Portuguese, it uses post-agreement spellings (ação, ótimo). For European Portuguese, it follows current Portuguese usage. The AI does not mix conventions between varieties within a single transcript.
Yes. European Portuguese reduces unstressed vowels so much that words like telefonema and despertar lose entire syllables acoustically. The AI is trained to recover these reduced sounds and produce complete, correctly spelled words.
Yes. The AI places clitic pronouns according to the detected variety: proclisis in Brazilian Portuguese (me diz, me conta, te amo) and enclisis in European Portuguese (diz-me, conta-me, amo-te), following the grammatical rules of each variety including mesoclisis in European Portuguese future tense (dir-lhe-ei).
Portuguese nasal vowels are contrastive and the AI distinguishes them reliably: mão (hand) vs mau (bad), lã (wool) vs lá (there), vim (I came) vs vi (I saw). The tilde, circumflex, and nasal consonant markers are placed correctly in the output.
Yes. The AI is trained on Brazilian speech from across the country, including the /r/ variation (retroflex in São Paulo, guttural in Rio, tapped in the South), the open/closed vowel differences, and Northeastern palatalization patterns. It produces standard Brazilian Portuguese spelling regardless of regional accent.
Transcribe Spanish audio and video with AI
Transcribe French audio and video with AI
Transcribe Italian audio and video with AI
Upload any audio file and get accurate text instantly
Upload audio in any language and translate it to 140+ languages
Upload audio or video and get ready-to-use subtitle files
Upload a file or paste a link from YouTube, podcasts, cloud storage, and 1,000+ platforms. Get an accurate transcript in minutes. No credit card required.