Galician audio to text: what to expect
Galician is a mid-resource language for the speech model. Clear speech gives a usable draft, but read it carefully. The typical mistake is a word written the Spanish or Portuguese way, because the model has much more data for both neighbours: hai might appear as hay, or xente as gente. Names and place names need checking as well.
Upload the recording as it is: MP3, M4A, WAV, OGG, OPUS, FLAC, a voice message or a phone recording. You can also record from your microphone in the upload box. Nothing needs converting first.
The transcript opens next to a player that highlights the current line. Click a line to hear it again, correct the word in place, and your changes save automatically.
Spelling: the official norm and its rivals
The output follows the official spelling set by the Real Academia Galega as closely as the model manages, with x for the sh sound (xente, xa) and ñ. Reintegrationist writers, who use spelling closer to Portuguese (lh, nh, ç), will need to convert the text by hand. Contractions such as co, coa, do and na, and the enclitic pronouns attached to verbs, are frequent spots for small errors.
Gheada and seseo, regional pronunciations of g and z/c, are common in parts of Galicia and can lead to misspelled words in the draft. The open and closed e and o that distinguish some Galician words are not marked in writing, so the model has to pick the word from context, and it sometimes picks the wrong one when two words differ only in that vowel.
Castrapo and Spanish in the same file
Many Galicians switch between Galician and Spanish, and everyday speech in cities often mixes the two (sometimes called castrapo). Choose by what dominates:
- Mostly Galician with Spanish words: keep Galician selected.
- Mostly Spanish with some Galician: choose Spanish.
- Avoid auto-detect for Galician; it frequently picks Spanish or Portuguese.
Who transcribes Galician and what next
Typical recordings are university lectures and theses, interviews for sociolinguistic or oral-history research, local radio and podcasts, council meetings, and conversations with older relatives in rural Galicia. Galician communities in Buenos Aires, Montevideo and Switzerland bring family recordings too. Teachers use transcripts of local speech as reading material, and journalists use them to pull quotes from long interviews quickly.
Export TXT, DOCX, PDF, CSV, SRT or VTT, with timestamps and speaker names switched on or off per export. Speaker labels are available on paid plans or with a Minute Pack.
For video files, see the Galician video to text page.