Loading datasets...
Task: ASR
Release Date: 6/17/2026
Format: MP3
Size: 79.21 GB
Share
A collection of read speech recordings in Catalan (català).
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.
ca)This datasheet is for cv-corpus-26.0-2026-06-12 of the Mozilla Common Voice Scripted Speech dataset for Catalan [català - ca]. The dataset contains 2716606 clips representing 3918.14 hours of recorded speech (3379.83 hours validated) from 36956 speakers, recorded from a text corpus of 1,326,608 sentences.
Catalan is a Romance language spoken by about 9 milion people mainly on the Mediterranean coast of the Iberian Peninsula.
It is an official language, along with Spanish or Castilian, in Catalonia, the Balearic Islands and the Valencian Community (where it is also called Valencian), while it is the only official language of the Principality of Andorra. It is also spoken, and has some administrative recognition, without reaching official status, in the eastern part of the autonomous community of Aragon, in the French department Pyrénées-Orientales (Eastern Pyrenees) and in the city of Alghero, on the island of Sardinia (Italy).
The language evolved from Vulgar Latin in the Middle Ages.
The main variants of Catalan are:
Central [ca-central]: It is the variant with the most speakers, as it encompasses the metropolitan area of Barcelona, extending to the region of Girona and the eastern half of Tarragona
Balearic [ca-balear]: The variant used in the Balearic Islands
Nord-Occidental [ca-nwestern]: Spoken in Andorra, Lleida and the western half of Tarragona in Catalonia, and the eastern part of Aragon
Septentrional [ca-northern]: Corresponds to the area of Roussillon and the northern part of Girona
Valencian: Spoken in the Valencian comunity, where it's also known as "Valencian"
Valencià meridional [ca-valencia-southern]
Alacantí [ca-valencia-alacant]
Valencià septentrional [ca-valencia-northern]
Tortosí [ca-valencia-tortosi]
Valencià central [ca-valencia-central]
Alguerese [ca-algueres]: Spoken in the city of Alghero, in Sardinia
| Code | Variant | Clips | Speakers |
|---|---|---|---|
| ca-central | Central | 1,421,260 (52.3%) | 6,429 (17.4%) |
| ca-nwestern | Nord-Occidental | 104,258 (3.8%) | 742 (2.0%) |
| ca-valencia-tortosi | Tortosí | 44,563 (1.6%) | 20 (0.1%) |
| ca-northern | Septentrional | 33,176 (1.2%) | 255 (0.7%) |
| ca-balear | Balear | 29,103 (1.1%) | 630 (1.7%) |
| ca-valencia-southern | Valencià meridional | 14,870 (0.5%) | 97 (0.3%) |
| ca-valencia-alacant | Alacantí | 7,311 (0.3%) | 73 (0.2%) |
| ca-valencia-northern | Valencià septentrional | 2,840 (0.1%) | 18 (0.0%) |
| ca-valencia-central | Valencià central | 973 (0.0%) | 54 (0.1%) |
| Code | Accent | Clips | Speakers |
|---|---|---|---|
| valencian | valencià | 112,074 (4.1%) | 699 (1.9%) |
| central | central | 100,957 (3.7%) | 736 (2.0%) |
| northwestern | nord-occidental | 49,185 (1.8%) | 121 (0.3%) |
| learner_other | aprenent (recent, des d'altres llengües) | 5,941 (0.2%) | 19 (0.1%) |
| balearic | balear | 3,645 (0.1%) | 76 (0.2%) |
| northern | septentrional | 1,677 (0.1%) | 49 (0.1%) |
| learner_es | aprenent (recent, des del castellà) | 1,322 (0.0%) | 54 (0.1%) |
| - | Other | 179,980 (6.6%) | 783 (2.1%) |
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 | 1,415,762 (52.1%) | 3,816 (10.3%) |
| female_feminine | Female, feminine | 567,213 (20.9%) | 3,348 (9.1%) |
| transgender | Transgender | - | - |
| non-binary | Non-binary | - | - |
| do_not_wish_to_say | Prefer not to say | 316 (0.0%) | 1 (0.0%) |
| - | Unspecified | 733,305 (27.0%) | 33,233 (89.9%) |
Gender declared: 1,983,301 of 2,716,606 clips (73.0%), 3,723 of 36,956 speakers (10.1%)
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 | 23,010 (0.8%) | 352 (1.0%) |
| twenties | Twenties | 106,326 (3.9%) | 1,133 (3.1%) |
| thirties | Thirties | 124,150 (4.6%) | 1,220 (3.3%) |
| fourties | Fourties | 319,320 (11.8%) | 1,858 (5.0%) |
| fifties | Fifties | 527,207 (19.4%) | 1,637 (4.4%) |
| sixties | Sixties | 785,195 (28.9%) | 922 (2.5%) |
| seventies | Seventies | 108,597 (4.0%) | 206 (0.6%) |
| eighties | Eighties | 5,909 (0.2%) | 10 (0.0%) |
| nineties | Nineties | 58 (0.0%) | 4 (0.0%) |
| - | Unspecified | 716,834 (26.4%) | 33,158 (89.7%) |
Age declared: 1,999,772 of 2,716,606 clips (73.6%), 3,798 of 36,956 speakers (10.3%)
Clip buckets
| Bucket | Clips |
|---|---|
| Validated | 2,343,375 (86.3%) |
| Invalidated | 144,402 (5.3%) |
| Other | 228,829 (8.4%) |
Training splits
| Split | Clips |
|---|---|
| Train | 1,218,545 (52.0%) |
| Dev | 16,416 (0.7%) |
| Test | 16,416 (0.7%) |
Training split coverage: 1,251,377 of 2,343,375 validated clips (53.4%)
The dataset contains 2343375 validated, 144402 invalidated, and 228829 unresolved clips. The average clip duration is 5.192 seconds.
Validated sentences: 1,310,098
| Category | Count |
|---|---|
| Unvalidated sentences | 16,510 |
| Pending sentences | 13,325 |
| Rejected sentences | 3,185 |
| Reported sentences | 10,122 |
The corpus contains 1,326,608 sentences: 1,310,098 validated and 16,510 unvalidated (13,325 pending review, 3,185 rejected), with 10,122 reported for review.
Catalan is written using the Latin alphabet (abcdefghijklmnopqrstuvwxyz), with the special characters ç and l·l. In addition, vowels can be accented (à, è, é, í, ò, ó, ú, ü, ï). The characters - (hyphen) and ' (apostrophe) are also part of Catalan orthography.
a b c ç d e f g h i j k l m n o p q r s t u v w x y z à è é í ò ó ú ï ü
There follows a randomly selected sample of five sentences from the corpus.
Les restes d'una torre i murs en un puig sobre la capella de Santa Anna.
Va ser membre fundador de l'Obra Cultural Balear.
La nostra experiència és molt bona, ja ho veurà.
La programació de Home Shopping Network va continuar durant el migdia, l'horari de màxima audiència i la matinada.
Al nivell superior hi ha motllura decorativa.
| Source | Sentences |
|---|---|
| wiki | 424,596 (32.4%) |
| covost2-ca | 262,235 (20.0%) |
| frases_agenda | 143,004 (10.9%) |
| new_catalan_cc0_corpus | 107,631 (8.2%) |
| new_sentences_from_catalan_newswire | 58,141 (4.4%) |
| Projecte Aina datasets sentences. CoQCat subset. | 57,588 (4.4%) |
| pccd | 33,674 (2.6%) |
| marius_serra_sentences | 32,114 (2.5%) |
| sentence-collector | 24,896 (1.9%) |
| Other | 165,676 (12.7%) |
| Code | Domain | Clips | Speakers |
|---|---|---|---|
| general | General | 2,041 (0.1%) | 106 (0.3%) |
| agriculture_food | Agriculture and Food | 75 (0.0%) | 20 (0.1%) |
| automotive_transport | Automotive and Transport | 41 (0.0%) | 16 (0.0%) |
| finance | Finance | 4 (0.0%) | 3 (0.0%) |
| service_retail | Service and Retail | 65 (0.0%) | 17 (0.0%) |
| healthcare | Healthcare | 42 (0.0%) | 20 (0.1%) |
| history_law_government | History, Law and Government | 71 (0.0%) | 25 (0.1%) |
| media_entertainment | Media and Entertainment | 53 (0.0%) | 14 (0.0%) |
| nature_environment | Nature and Environment | 155 (0.0%) | 25 (0.1%) |
| news_current_affairs | News and Current Affairs | 45 (0.0%) | 19 (0.1%) |
| technology_robotics | Technology and Robotics | 38 (0.0%) | 19 (0.1%) |
| language_fundamentals | Language Fundamentals | 45 (0.0%) | 17 (0.0%) |
It is recommended to normalize instances of the geminate L, which can take the equivalent forms of l·l or ŀl.
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
sentence - the sentence to be read aloud
sentence_id - unique identifier for the sentence
sentence_domain - domain classification(s) of the sentence
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
locale - locale code of the language
segment - if sentence belongs to a custom dataset segment, it will be listed here
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)
Carme Armentano
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