Loading datasets...
Task: ASR
Release Date: 3/23/2026
Format: MP3
Size: 7.81 GB
Share
A collection of read speech recordings in Galician (Galego).
Restrictions/Special Constraints
None provided.
Forbidden Usage
It is forbidden to attempt to determine the identity of speakers in the Common Voice datasets. It is forbidden to re-host or re-share this dataset.
Intended Use
This dataset is intended to be used for training and evaluating automatic speech recognition (ASR) models. It may also be used for applications relating to computer-aided language learning (CALL) and language or heritage revitalisation.
gl)This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for Galician [Galego - gl]. The dataset contains 288450 clips representing 399.69 hours of recorded speech (308 hours validated) from 6592 speakers, recorded from a text corpus of 1,005,836 sentences.
Galician is the language of Galicia (Spain). In addition to being a language of use in this territory, it is spoken in other bordering areas, although without official status (Asturias, León, Zamora). It belongs to the Indo-European language family, specifically the Romance languages, along with Spanish, Portuguese and Catalan. It also has a close relationship with Portuguese since they share historical origins, but today they constitute two separate languages with each one having its own respective standard. The state of vitality of the language is “Institutional” according to Ethnologue https://www.ethnologue.com/language/glg/, that is, the language is developed, has a standard variety, is taught at all stages of the Galician educational system and has institutional support. However, from the point of view of sociolinguistics, both the autonomous community of Galicia and the other territories that speak the Galician language present typical features of a situation of linguistic conflict (diglossia), resulting from a historical process of replacement of Galician (a minority language) by Spanish.
The Galician language is a rich and diverse reality, and its linguistic variation is manifested in spontaneous speech. Thus, aspects such as age, geography, sociocultural level or communicative context also influence linguistic uses. At the dialectal level, Galician presents differentiating features between the eastern and western areas, as well as between the north and the south: phonetic changes (such as the use of gheada or seseo), lexical variations (different words to designate the same object) or even small morpho-syntactic differences.
| Code | Accent | Clips | Speakers |
|---|---|---|---|
| normativo | Normativo (estándar) | 64,917 (22.5%) | 1,633 (24.8%) |
| central | Central (gheada) | 29,691 (10.3%) | 394 (6.0%) |
| atlantico | Atlántico (seseo e gheada) | 24,693 (8.6%) | 500 (7.6%) |
| neofalante | Neofalante | 13,038 (4.5%) | 367 (5.6%) |
| oriental | Oriental (común en zona oriental) | 11,902 (4.1%) | 374 (5.7%) |
| - | Other | 7,954 (2.8%) | 194 (2.9%) |
The dataset includes the following self-declared age and gender distributions. A coverage summary is shown below each table.
Self-declared gender information. The table shows clip and speaker counts with percentages. Speakers who did not declare a gender are listed as Unspecified. A dash (-) indicates zero.
| Code | Gender | Clips | Speakers |
|---|---|---|---|
| male_masculine | Male, masculine | 34,848 (12.1%) | 363 (5.5%) |
| female_feminine | Female, feminine | 68,461 (23.7%) | 712 (10.8%) |
| transgender | Transgender | - | - |
| non-binary | Non-binary | 25 (0.0%) | 3 (0.0%) |
| do_not_wish_to_say | Prefer not to say | 1,663 (0.6%) | 3 (0.0%) |
| - | Unspecified | 183,453 (63.6%) | 5,823 (88.3%) |
Gender declared: 104,997 of 288,450 clips (36.4%), 769 of 6,592 speakers (11.7%)
Self-declared age information. The table shows clip and speaker counts with percentages. Speakers who did not declare an age are listed as Unspecified. A dash (-) indicates zero.
| Code | Age | Clips | Speakers |
|---|---|---|---|
| teens | Teens | 10,559 (3.7%) | 112 (1.7%) |
| twenties | Twenties | 51,195 (17.7%) | 502 (7.6%) |
| thirties | Thirties | 18,791 (6.5%) | 239 (3.6%) |
| fourties | Fourties | 25,398 (8.8%) | 233 (3.5%) |
| fifties | Fifties | 31,602 (11.0%) | 150 (2.3%) |
| sixties | Sixties | 2,512 (0.9%) | 34 (0.5%) |
| seventies | Seventies | 130 (0.0%) | 4 (0.1%) |
| eighties | Eighties | - | - |
| nineties | Nineties | - | - |
| - | Unspecified | 148,263 (51.4%) | 5,706 (86.6%) |
Age declared: 140,187 of 288,450 clips (48.6%), 886 of 6,592 speakers (13.4%)
Clip buckets
| Bucket | Clips |
|---|---|
| Validated | 222,281 (77.1%) |
| Invalidated | 34,523 (12.0%) |
| Other | 31,646 (11.0%) |
Training splits
| Split | Clips |
|---|---|
| Train | 184,932 (83.2%) |
| Dev | 15,255 (6.9%) |
| Test | 15,266 (6.9%) |
Training split coverage: 215,453 of 222,281 validated clips (96.9%)
The dataset contains 222281 validated, 34523 invalidated, and 31646 unresolved clips. The average clip duration is 4.988 seconds.
Validated sentences: 1,002,019
| Category | Count |
|---|---|
| Unvalidated sentences | 3,817 |
| Pending sentences | 3,253 |
| Rejected sentences | 564 |
| Reported sentences | 946 |
The corpus contains 1,005,836 sentences: 1,002,019 validated and 3,817 unvalidated (3,253 pending review, 564 rejected), with 946 reported for review.
The writing system of Galician is based on the Latin alphabet, like other Romance languages (Spanish, Portuguese or French). Its orthography is regulated by the corresponding regional institutions that establish the official rules of use (NOMIG).
Galician uses the 23 letters of the basic Latin alphabet: 5 vowels (a, e, i, o, u) and 18 consonants. The vowel system of Galician is composed of seven oral vowels (/a/, /e/, /ɛ/, /i/, /o/, /ɔ/ and /u/), as the mid vowels ("e" and "o") have two possible pronunciations: open (/ɛ/, /ɔ/) and closed (/e/, /o/). At the consonant level, the letters Ç, J, K, W and Y are supported in foreign words, proper names, abbreviations and international acronyms. The Galician alphabet also has numerous digraphs: combinations of two letters that represent a single sound: "ch", "gu", "ll", "nh", "qu", "rr". In addition, the presence of the phoneme /ʃ/ represented by the letter “x”, the fricative sound /ʒ/, the distinction between /s/ and /θ/ which depends on the dialect areas, the gheada, etc., should be highlighted in the Galician consonant system.
Accentuation in Galician is quite regular and follows clear rules. A graphic "diacritical accent" is also used to distinguish pairs of words with the same spelling but different meanings, e.g., bóla (sphere, ball), bola (round type of bread).
a, á, b, c, d, e, é, f, g, h, i, í, ï, l, m, n, o, ó, p, q, r, s, t, u, ú, ü, v, x, z
There follows a randomly selected sample of five sentences from the corpus.
"A finalidade, ""tentar atopar unha solución""."
Os primeiros minutos do partido, tolos, xa anticipaban un resultado incerto.
Din que lles levaron máis da metade da froita dalgunhas das variedades.
Fotógrafo coñecido co pseudónimo de Robert Capa.
En canto ás Sueiras, o empregado parecía inseguro.
| Source | Sentences |
|---|---|
| TVG (2019-2022). CC0 Waiver signed. | 166,845 (16.7%) |
| Parlamento de Galicia (2015-2022). CC0 Waiver signed. | 166,157 (16.6%) |
| Nós Diario (2012-2022). CC0 Waiver signed. | 140,026 (14.0%) |
| wiki | 92,840 (9.3%) |
| Praza.gal (2012-2022). CC0 Waiver signed. | 89,340 (8.9%) |
| Diario dixital Que pasa na Costa | 86,655 (8.6%) |
| Consello da Cultura Galega / Gran Enciclopedia Galega | 84,104 (8.4%) |
| sentence-collector | 32,593 (3.3%) |
| Editorial Galaxia | 28,948 (2.9%) |
| Other | 114,511 (11.4%) |
| Code | Domain | Clips | Speakers |
|---|---|---|---|
| general | General | 269 (0.1%) | 117 (1.8%) |
| agriculture_food | Agriculture and Food | 34 (0.0%) | 24 (0.4%) |
| automotive_transport | Automotive and Transport | 4 (0.0%) | 3 (0.0%) |
| finance | Finance | 40 (0.0%) | 27 (0.4%) |
| service_retail | Service and Retail | 17 (0.0%) | 17 (0.3%) |
| healthcare | Healthcare | 3 (0.0%) | 3 (0.0%) |
| history_law_government | History, Law and Government | 52 (0.0%) | 37 (0.6%) |
| media_entertainment | Media and Entertainment | 41 (0.0%) | 32 (0.5%) |
| nature_environment | Nature and Environment | 15 (0.0%) | 13 (0.2%) |
| news_current_affairs | News and Current Affairs | 14 (0.0%) | 13 (0.2%) |
| technology_robotics | Technology and Robotics | 5 (0.0%) | 5 (0.1%) |
| language_fundamentals | Language Fundamentals | 19 (0.0%) | 14 (0.2%) |
Each row of a tsv file represents a single audio clip, and contains the following information:
client_id - hashed UUID of a given user
path - relative path of the audio file
text - supposed transcription of the audio
up_votes - number of people who said audio matches the text
down_votes - number of people who said audio does not match text
age - age of the speaker1
gender - gender of the speaker1
accents - accents of the speaker1
variant - variant of the language1
segment - if sentence belongs to a custom dataset segment, it will be listed here
prompt_upvotes - number of upvotes the sentence prompt received
prompt_reports - number of reports the sentence prompt received
is_edited - whether the clip's transcription has been edited
validated_sentences.tsvThe validated_sentences.tsv file contains one row per validated sentence in the text corpus:
sentence_id - unique identifier for the sentence
sentence - the sentence text
variant - the variant of the language
sentence_domain - the domain(s) the sentence belongs to
source - the source the sentence was collected from
is_used - whether the sentence is still in circulation for recording
clips_count - number of clips recorded for this sentence
unvalidated_sentences.tsvThe unvalidated_sentences.tsv file contains one row per unvalidated sentence in the text corpus:
sentence_id - unique identifier for the sentence
sentence - the sentence text
variant - the variant of the language
sentence_domain - the domain(s) the sentence belongs to
source - the source the sentence was collected from
up_votes - number of upvotes the sentence received
down_votes - number of downvotes the sentence received
status - current status of the sentence (pending or rejected)
This dataset is released under the Creative Commons Zero (CC-0) licence. By downloading this data you agree to not determine the identity of speakers in the dataset.
For a full list of age, gender, and accent options, see the demographics spec. These will only be reported if the speaker opted in to provide that information. ↩ ↩2 ↩3 ↩4