Skip to content

Datasets

Hugging Face datasets with plain-language descriptions: what's inside and what models to train on them. 4,788 in the catalog so far, with new ones added every hour.

Data type

For which models

Find a dataset for your goal

Answered by Gemini 4 · the chat is kept only in this tab

Datasetdisco-eth/WorldSpeechTop pick

WorldSpeech: 65k hours of multilingual speech

A multilingual ASR corpus with 65k hours of human transcripts for 127 language variants.

  • Automatic speech recognition
  • Text → speech
  • Audio classification
  • Afrikaans
  • Amharic
  • Arabic
  • +9
34.1K 7710M–100M rows
Datasetkapturecx/bolAIndia

Multilingual Indian speech and dialect corpus

Large-scale collection of audio recordings, texts, and documents in rare Indian languages and dialects for ASR systems.

  • Automatic speech recognition
  • Angika
  • Assamese
  • Awadhi
  • +9
34.1K 710M–100M rows🔒 on request
Datasetgoaicorp/mos-speech-internal

Private Moore Speech Corpus (GO AI Corp)

A large internal speech corpus in the Moore language with French translations and metadata for ASR and TTS tasks.

  • Automatic speech recognition
  • Text → speech
  • Mossi
  • French
33.1K 0100K–1M rows🔒 on request
Datasetfsicoli/common_voice_17_0

Mozilla Common Voice 17.0 speech corpus mirror

Large multilingual audio dataset for speech recognition converted from Mozilla Common Voice version 17.

  • Automatic speech recognition
  • Abkhazian
  • Afrikaans
  • Amharic
  • +9
32.1K 19100B–1T rows
Datasetopenbhasha/rsml-bench

Telugu transcribed speech with disfluencies and tags

Nearly 16 hours of Telugu audio with detailed transcription, timecodes, and speech feature markup.

  • Automatic speech recognition
  • Telugu
31.7K 0up to 1K rows🔒 on request
Datasetfsicoli/common_voice_22_0

Common Voice 22.0 multilingual speech corpus

Unofficial export of the 22nd version of Mozilla's open speech dataset for speech recognition tasks.

  • Automatic speech recognition
  • Abkhazian
  • Afrikaans
  • Amharic
  • +9
31.7K 20100B–1T rows
DatasetHebrish/Irish-English-Sentences

Irish English accented dataset

Audio recordings of everyday sentences in Irish English.

  • Automatic speech recognition
  • English
31K 0up to 1K rows🔒 on request
Datasetibarbelo/sales-conversations-sample

Real conversational audio data sample

A small sample of two-way conversations with separated speaker tracks for ASR.

  • Automatic speech recognition
  • Audio classification
  • English
30.4K 1up to 1K rows🔒 on request
Datasetgoogle/WaxalNLPTop pick

Waxal: multilingual African speech corpus

Audio and text for ASR and TTS across dozens of African languages—a large open speech corpus from Google.

  • Automatic speech recognition
  • Text → speech
  • Acoli
  • Akan
  • Amharic
  • +9
28.1K 3131M–10M rows
DatasetMuncy/AudiobookRuTop pick

Russian audiobook chapters with metadata

Around 893,000 chapters from 96,000 Russian audiobooks with audio files and descriptions.

  • Automatic speech recognition
  • Text → speech
  • Audio classification
  • Russian
27.2K 8100K–1M rows
Datasetchantysothy/khmer-tipitaka-buddha-life-speech

Khmer speech: Tipitaka and Buddha life

Read Khmer speech based on Tipitaka texts for ASR and TTS tasks with manual access moderation.

  • Automatic speech recognition
  • Text → speech
  • Khmer
27.1K 0100K–1M rows🔒 on request
DatasetMC7ever/hf-training-corpus

Multimodal HF Corpus of 92K Datasets

Large-scale extraction of texts, images, audio, and video from open datasets in a unified parquet format.

  • Text generation
  • Image → text
  • Automatic speech recognition
  • English
26.8K 1100K–1M rows
Datasetdavidscripka/MIT_environmental_impulse_responses

MIT real-world room impulse responses

271 audio files featuring acoustic impulse responses for speech augmentation and reverberation simulation.

  • Audio classification
  • Automatic speech recognition
25.9K 9up to 1K rows
Datasetbond005/sberdevices_golos_10h_crowd

SberDevices Golos Russian speech corpus subset (crowd)

16 kHz Russian speech audio recordings with manual annotations from the crowdsourced part of the Golos corpus for ASR tasks.

  • Automatic speech recognition
  • Audio classification
  • Russian
25.8K 610K–100K rows
Datasetspeechbrain/LoquaciousSetTop pick

25,000 hours of English speech with transcripts

A large heterogeneous English speech corpus with transcriptions and ready-made subsets for training ASR models.

  • Automatic speech recognition
  • English
25.7K 6410M–100M rows
Datasetfsicoli/common_voice_16_0

Multilingual Common Voice 16.0 speech corpus

An unofficial copy of the open Mozilla Common Voice 16.0 speech corpus for ASR tasks.

  • Automatic speech recognition
  • Abkhazian
  • Afrikaans
  • Amharic
  • +9
25.1K 4100B–1T rows
Datasetaoxo/t2a-mommy

English female ASMR audio corpus with word timestamps

Female ASMR voice recordings in AAC format with Whisper word-level alignment and non-speech sound tags for ASR and generation tasks.

  • Audio classification
  • Automatic speech recognition
  • English
24.5K 1
Datasetparler-tts/mls_eng

LibriVox English audio speech for ASR and TTS

Large English audiobook speech and text corpus (10m to 100m records) optimized for streaming.

  • Automatic speech recognition
  • Text → speech
  • Text → audio
  • English
23.9K 4110M–100M rows
Datasettwangodev/librivox-mirrorTop pick

Multilingual LibriVox Audiobooks Mirror

Over 133,000 hours of voiced books across 86 languages for training speech recognition and synthesis systems.

  • Automatic speech recognition
  • Text → speech
  • Multilingual
22.9K 0100K–1M rows
Datasetlaubonghaudoi/legco-speechTop pick

Hong Kong Legislative Council speech audio data

A large-scale audio dataset and transcripts of Legislative Council meetings in Cantonese, featuring raw and segmented recordings.

  • Automatic speech recognition
  • Audio → audio
  • Audio classification
  • Cantonese
22.9K 31M–10M rows
Datasetchikingsley/l2-arctic-release-v5.0

L2-ARCTIC accented English speech with phonetic labels

Speech corpus of 26,867 audio readings by 24 non-native English speakers with manual phoneme labeling for ASR and pronunciation analysis.

  • Automatic speech recognition
  • Audio classification
  • English
22.4K 110K–100K rows
Datasetlab260/balalaika-sft-rover7-punctTop pick

Large Russian speech corpus for ASR with punctuation

Nearly 12,000 hours of Russian speech with consensus transcriptions from seven ASR models and added punctuation.

  • Automatic speech recognition
  • Russian
21.9K 31M–10M rows
DatasetMuno459/QuranLab-Arabic-Speech

Arabic speech dataset for ASR

Wideband Arabic speech in segments for training ASR and TTS models.

  • Automatic speech recognition
  • Text → speech
  • Arabic
21.8K 110M–100M rows🔒 on request