Southeast Asian speech data in 6 languages
Spontaneous two-person conversation in Indonesian, Malay, Thai, Vietnamese, Tagalog and Singapore English: 12,000 hours from 1,200 native speakers recorded in Indonesia, Malaysia, Thailand, Vietnam, the Philippines and Singapore, at $60 to $75 per hour of audio.
Compare the Southeast Asian datasets
| Language | Recorded in | Hours | Speakers | Per hour |
|---|---|---|---|---|
| Indonesian | Indonesia | 2,000 | 200 | $65 |
| Malay | Malaysia | 2,000 | 200 | $65 |
| Thai | Thailand | 2,000 | 200 | $75 |
| Vietnamese | Vietnam | 2,000 | 200 | $65 |
| Tagalog | Philippines | 2,000 | 200 | $65 |
| Singapore English | Singapore | 2,000 | 200 | $60 |
| All 6 | 6 countries | 12,000 | 1,200 | $60 to $75 |
List prices in US dollars per hour of audio. Need another language, more hours or your own recording setup? We collect speech data to spec.
What differs between these languages
Indonesian
Formal and everyday Indonesian are quite different. People drop prefixes, use Jakarta slang like "gue", "lo" and "nggak", and mix in regional languages such as Javanese or Sundanese.
Malay
Malaysian Malay has strong regional accents (Kelantan and Terengganu speech is hard for outsiders to follow), and speakers mix in English constantly.
Thai
Thai has five tones that change meaning, long and short vowels, and a script that writes words without spaces.
Vietnamese
Vietnamese is tonal, with six tones in the northern standard and fewer in the south, and each tone is written as a mark on the vowel.
Tagalog
Everyday Tagalog in Manila is usually Taglish, a constant mix of Tagalog and English inside the same sentence, plus Spanish loanwords.
Singapore English
Singapore English runs from standard English to colloquial Singlish, with particles like "lah", "leh" and "lor" and words from Hokkien, Malay and Tamil.
What every dataset includes
| Speech | Spontaneous two-person conversations recorded in-country, not read scripts |
|---|---|
| Audio | Dual-channel, one channel per speaker |
| Transcripts | Time-aligned and human-verified |
| Metadata | Age, gender, region and recording device for every speaker |
| Consent | Every speaker signed explicit commercial-use consent; the full consent chain is provided for legal review |
| License | Commercial use, including training and evaluating ASR, TTS and other voice AI models |
| Price | $60 to $75 per hour of audio |
Built for
Training and evaluating speech models for Southeast Asian markets:
- Automatic speech recognition (ASR / STT)
- Text-to-speech & voice synthesis prosody
- Voice agents and IVR
- Speaker diarization & turn-taking models
- Speech LLM / audio-understanding fine-tuning
Frequently asked questions
Is Malay speech data the same as Indonesian?
No. The two languages are close, but spelling and vocabulary differ, so Indonesian data doesn't transfer cleanly to a Malay model. The Malay dataset is recorded in Malaysia (2,000 hours) and the Indonesian dataset in Indonesia (2,000 hours).
Can I license several Southeast Asian languages together?
Yes. Each dataset is licensed at its own published price, with volume and multi-language discounts on request. All 6 share one format: dual-channel audio of spontaneous two-person conversation, time-aligned and human-verified transcripts, and the same speaker metadata, so they combine into one multilingual training set.
How much does Southeast Asian speech data cost?
List prices are per hour of audio: $60 for Singapore English; $65 for Indonesian, Malay, Vietnamese and Tagalog; $75 for Thai. Each dataset page shows its price, and volume and multi-language discounts are available on request.
Can you record other Southeast Asian languages?
Yes. Other languages, such as Khmer, Burmese, Lao or Javanese, can be recorded to spec by native speakers, scripted or spontaneous, with time-aligned transcripts and documented consent. We confirm feasibility and send a quote within one business day.
Hear the languages before you buy
Tell us which languages you need and share your company email. We’ll reach out shortly with representative audio, matching transcripts and the metadata schema, so you can check quality before you commit.