Loading datasets...
Task: ASR
Release Date: 6/17/2026
Format: MP3
Size: 88.14 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-26.0-2026-06-12 of the Mozilla Common Voice Scripted Speech dataset for English [English - en]. The dataset contains 2583051 clips representing 3780.95 hours of recorded speech (2784.88 hours validated) from 100172 speakers, recorded from a text corpus of 1,721,897 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 | 573,393 (22.2%) | 10,970 (11.0%) |
| england | England English | 204,106 (7.9%) | 3,445 (3.4%) |
| indian | India and South Asia (India, Pakistan, Sri Lanka) | 152,723 (5.9%) | 3,029 (3.0%) |
| canada | Canadian English | 101,061 (3.9%) | 1,220 (1.2%) |
| australia | Australian English | 69,363 (2.7%) | 948 (0.9%) |
| scotland | Scottish English | 68,131 (2.6%) | 270 (0.3%) |
| african | Southern African (South Africa, Zimbabwe, Namibia) | 60,292 (2.3%) | 441 (0.4%) |
| newzealand | New Zealand English | 20,768 (0.8%) | 230 (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 | 4,722 (0.2%) | 112 (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) | 1,231 (0.0%) | 74 (0.1%) |
| southatlandtic | South Atlantic (Falkland Islands, Saint Helena) | 332 (0.0%) | 9 (0.0%) |
| other | Other | 200,693 (7.8%) | 1,681 (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 | 1,117,432 (43.3%) | 18,517 (18.5%) |
| female_feminine | Female, feminine | 457,904 (17.7%) | 5,563 (5.6%) |
| transgender | Transgender | 156 (0.0%) | 11 (0.0%) |
| non-binary | Non-binary | 411 (0.0%) | 17 (0.0%) |
| do_not_wish_to_say | Prefer not to say | 2,009 (0.1%) | 44 (0.0%) |
| - | Unspecified | 1,004,989 (38.9%) | 78,307 (78.2%) |
Gender declared: 1,578,062 of 2,583,051 clips (61.1%), 21,865 of 100,172 speakers (21.8%)
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 | 151,296 (5.9%) | 3,255 (3.2%) |
| twenties | Twenties | 641,400 (24.8%) | 11,552 (11.5%) |
| thirties | Thirties | 357,504 (13.8%) | 5,369 (5.4%) |
| fourties | Fourties | 240,835 (9.3%) | 2,651 (2.6%) |
| fifties | Fifties | 136,070 (5.3%) | 1,590 (1.6%) |
| sixties | Sixties | 115,750 (4.5%) | 915 (0.9%) |
| seventies | Seventies | 17,522 (0.7%) | 365 (0.4%) |
| eighties | Eighties | 2,427 (0.1%) | 55 (0.1%) |
| nineties | Nineties | 306 (0.0%) | 12 (0.0%) |
| - | Unspecified | 919,941 (35.6%) | 76,964 (76.8%) |
Age declared: 1,663,110 of 2,583,051 clips (64.4%), 23,208 of 100,172 speakers (23.2%)
Clip buckets
| Bucket | Clips |
|---|---|
| Validated | 1,902,564 (73.7%) |
| Invalidated | 313,382 (12.1%) |
| Other | 367,105 (14.2%) |
Training splits
| Split | Clips |
|---|---|
| Train | 1,147,812 (60.3%) |
| Dev | 16,403 (0.9%) |
| Test | 16,403 (0.9%) |
Training split coverage: 1,180,618 of 1,902,564 validated clips (62.1%)
The dataset contains 1902564 validated, 313382 invalidated, and 367105 unresolved clips. The average clip duration is 5.27 seconds.
Validated sentences: 1,681,666
| Category | Count |
|---|---|
| Unvalidated sentences | 40,231 |
| Pending sentences | 36,278 |
| Rejected sentences | 3,953 |
| Reported sentences | 9,773 |
The corpus contains 1,721,897 sentences: 1,681,666 validated and 40,231 unvalidated (36,278 pending review, 3,953 rejected), with 9,773 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.
For example, a heavily rhythmic speech filled with mnemonic devices enhances memory and recall.
Opposition to construction came from various citizens' groups and different levels of local government.
"""Perfect Balance"" was produced and engineered by Lionel Hicks."
It provides people with structure and purpose and a sense of identity.
I learned this from the late Mr. W. Simpson.
| Source | Sentences |
|---|---|
| wiki | 1,537,302 (92.9%) |
| sentence-collector | 61,569 (3.7%) |
| covost2-xx_en | 28,881 (1.7%) |
| Other | 26,425 (1.6%) |
| Code | Domain | Clips | Speakers |
|---|---|---|---|
| general | General | 677 (0.0%) | 334 (0.3%) |
| agriculture_food | Agriculture and Food | 169 (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 | 126 (0.0%) | 94 (0.1%) |
| media_entertainment | Media and Entertainment | 118 (0.0%) | 92 (0.1%) |
| nature_environment | Nature and Environment | 64 (0.0%) | 46 (0.0%) |
| news_current_affairs | News and Current Affairs | 13 (0.0%) | 13 (0.0%) |
| technology_robotics | Technology and Robotics | 104 (0.0%) | 75 (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
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)
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