100 languages for AI training data, with deep Indian coverage
Every language below has its own page covering speech, text, conversational and translation data in that language. This directory covers the most widely spoken languages worldwide, grouped by region, plus the deeper Indian language coverage described on our Indian language data page. Don't see yours? Ask us, since contributors are matched to your brief, not pulled from a fixed catalog.
We also hold around 35,000 hours of off-the-shelf speech, prompts and conversations already collected across our network. Ask whether your language is already covered before commissioning a fresh collection.
11 languages from East and Southeast Asia
East Asian and Southeast Asian languages, each with their own writing system and speech patterns.
25 languages from Europe
Major European languages, each with regional accents worth naming in your brief.
4 languages from the Americas
Indigenous and creole languages spoken across Latin America and the Caribbean.
Questions about language coverage
The datasets FAQ covers questions that apply across every category.
Is this the full list of languages you support?
This directory lists 100 languages, each with its own dedicated page. It covers the most widely spoken languages worldwide plus deeper coverage within India. If your language isn't listed, ask us, since we source contributors to the brief rather than from a fixed catalog.
Do you have deeper coverage of Indian languages than this list shows?
Yes. See Indian language data for that depth, including languages beyond the ones listed here.
Can a project combine more than one language from this list?
Yes. Many engagements mix languages, domains and data types under one brief and one named contact.
Do you have any off-the-shelf data already collected?
Yes. We hold around 35,000 hours of off-the-shelf speech, prompts and conversations already collected across our network, and 71 languages have their own ready-to-license dataset, listed on our OTS speech datasets page. If yours isn't listed, ask: checking costs nothing before you commission a fresh collection.
Request a language
Pick the data type, name the language and a rough volume, and send it. We'll produce a small sample free before you commit.
- Your named contact replies within one business day.
- We sign an NDA before any data changes hands.
- We produce your free sample, then agree the full scope.
This opens your email app with the brief filled in. Nothing is sent until you press send. You can also use the contact page.