Loading datasets...
License:
CC0-1.0
Steward:
Common Voice
Task: ASR
Release Date: 3/30/2026
Format: MP3
Size: 87.84 GB
Share
A collection of read speech recordings in English (English).
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.
en)This datasheet is for cv-corpus-25.0-2026-03-09 of the Mozilla Common Voice Scripted Speech dataset for English [English - en]. The dataset contains 2574404 clips representing 3765.76 hours of recorded speech (2751.56 hours validated) from 99724 speakers, recorded from a text corpus of 1,720,488 sentences.
English is a West Germanic language with origins in England. There are an estimated 1.5 billion English speakers, making it the most widely spoken language in the world. English is commonly learned as a second language in many countries.
| Code | Accent | Clips | Speakers |
|---|---|---|---|
| us | United States English | 572,552 (22.2%) | 10,918 (10.9%) |
| england | England English | 203,338 (7.9%) | 3,419 (3.4%) |
| indian | India and South Asia (India, Pakistan, Sri Lanka) | 152,535 (5.9%) | 3,011 (3.0%) |
| canada | Canadian English | 98,459 (3.8%) | 1,214 (1.2%) |
| australia | Australian English | 69,328 (2.7%) | 941 (0.9%) |
| scotland | Scottish English | 68,091 (2.6%) | 267 (0.3%) |
| african | Southern African (South Africa, Zimbabwe, Namibia) | 60,057 (2.3%) | 429 (0.4%) |
| newzealand | New Zealand English | 20,763 (0.8%) | 229 (0.2%) |
| ireland | Irish English | 11,204 (0.4%) | 266 (0.3%) |
| philippines | Filipino | 7,492 (0.3%) | 207 (0.2%) |
| hongkong | Hong Kong English | 7,004 (0.3%) | 190 (0.2%) |
| singapore | Singaporean English | 4,712 (0.2%) | 110 (0.1%) |
| malaysia | Malaysian English | 4,235 (0.2%) | 154 (0.2%) |
| wales | Welsh English | 3,032 (0.1%) | 119 (0.1%) |
| bermuda | West Indies and Bermuda (Bahamas, Bermuda, Jamaica, Trinidad) | 1,161 (0.0%) | 74 (0.1%) |
| southatlandtic | South Atlantic (Falkland Islands, Saint Helena) | 332 (0.0%) | 9 (0.0%) |
| other | Other | 199,478 (7.7%) | 1,651 (1.7%) |
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 | 1,114,005 (43.3%) | 18,513 (18.6%) |
| female_feminine | Female, feminine | 457,042 (17.8%) | 5,531 (5.5%) |
| transgender | Transgender | 156 (0.0%) | 11 (0.0%) |
| non-binary | Non-binary | 401 (0.0%) | 17 (0.0%) |
| do_not_wish_to_say | Prefer not to say | 1,989 (0.1%) | 42 (0.0%) |
| - | Unspecified | 1,000,666 (38.9%) | 77,894 (78.1%) |
Gender declared: 1,573,738 of 2,574,404 clips (61.1%), 21,830 of 99,724 speakers (21.9%)
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 | 151,017 (5.9%) | 3,237 (3.2%) |
| twenties | Twenties | 639,952 (24.9%) | 11,493 (11.5%) |
| thirties | Thirties | 357,223 (13.9%) | 5,348 (5.4%) |
| fourties | Fourties | 240,317 (9.3%) | 2,631 (2.6%) |
| fifties | Fifties | 133,561 (5.2%) | 1,584 (1.6%) |
| sixties | Sixties | 114,052 (4.4%) | 910 (0.9%) |
| seventies | Seventies | 17,522 (0.7%) | 365 (0.4%) |
| eighties | Eighties | 2,427 (0.1%) | 55 (0.1%) |
| nineties | Nineties | 305 (0.0%) | 11 (0.0%) |
| - | Unspecified | 918,028 (35.7%) | 76,628 (76.8%) |
Age declared: 1,656,376 of 2,574,404 clips (64.3%), 23,096 of 99,724 speakers (23.2%)
Clip buckets
| Bucket | Clips |
|---|---|
| Validated | 1,881,067 (73.1%) |
| Invalidated | 310,527 (12.1%) |
| Other | 382,810 (14.9%) |
Training splits
| Split | Clips |
|---|---|
| Train | 1,147,812 (61.0%) |
| Dev | 16,403 (0.9%) |
| Test | 16,403 (0.9%) |
Training split coverage: 1,180,618 of 1,881,067 validated clips (62.8%)
The dataset contains 1881067 validated, 310527 invalidated, and 382810 unresolved clips. The average clip duration is 5.266 seconds.
Validated sentences: 1,681,626
| Category | Count |
|---|---|
| Unvalidated sentences | 38,862 |
| Pending sentences | 34,938 |
| Rejected sentences | 3,924 |
| Reported sentences | 9,638 |
The corpus contains 1,720,488 sentences: 1,681,626 validated and 38,862 unvalidated (34,938 pending review, 3,924 rejected), with 9,638 reported for review.
The English writing system is based off of the latin alphabet.
a b c d e f g h i j k l m n o p q r s t u v w x y z
There follows a randomly selected sample of five sentences from the corpus.
Pus collects in the renal pelvis and causes distension of the kidney.
The Rockingham site was redeveloped and sub-divided into residential lots.
One form, the anoscope, resembles a tube that has a removable bullet-shaped insert.
The following four referee pairs were selected.
The show also featured Mike Patton and Fred Frith as guests.
| Source | Sentences |
|---|---|
| wiki | 1,537,302 (92.9%) |
| sentence-collector | 61,569 (3.7%) |
| covost2-xx_en | 28,881 (1.7%) |
| Other | 26,385 (1.6%) |
| Code | Domain | Clips | Speakers |
|---|---|---|---|
| general | General | 660 (0.0%) | 324 (0.3%) |
| agriculture_food | Agriculture and Food | 166 (0.0%) | 115 (0.1%) |
| automotive_transport | Automotive and Transport | 8 (0.0%) | 7 (0.0%) |
| finance | Finance | 44 (0.0%) | 33 (0.0%) |
| service_retail | Service and Retail | 31 (0.0%) | 24 (0.0%) |
| healthcare | Healthcare | 26 (0.0%) | 22 (0.0%) |
| history_law_government | History, Law and Government | 123 (0.0%) | 94 (0.1%) |
| media_entertainment | Media and Entertainment | 117 (0.0%) | 91 (0.1%) |
| nature_environment | Nature and Environment | 61 (0.0%) | 44 (0.0%) |
| news_current_affairs | News and Current Affairs | 13 (0.0%) | 13 (0.0%) |
| technology_robotics | Technology and Robotics | 101 (0.0%) | 72 (0.1%) |
| language_fundamentals | Language Fundamentals | 11 (0.0%) | 10 (0.0%) |
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