Translate
Read source text and write a translation in your language.
Open for contributors ยท Every language welcome
Your voice builds languages that last.
Native and fluent speakers build open datasets for 7,000+ languages worldwide. Translate, transcribe, record, and peer-review, one contribution at a time.
- 7,000+
- languages worldwide
- 5 task types
- to contribute
- 100%
- open data
Most of the world's languages are invisible to AI.
Over 7,000 languages are spoken worldwide. Only a fraction have enough recorded data to train a speech model, build a translator, or power a voice assistant. The rest are excluded by default, not because they're less valuable, but because no one has gathered the data.
All Voices exists to close that gap. Native and fluent speakers contribute translations, recordings, and image descriptions through a mobile app. Every submission is peer-reviewed, quality-checked, and exported as open research data.
7,000+languages worldwide
How it works
Five ways to contribute.
Pick any task, in any language you speak. Each contribution is reviewed by peers and earns points.
Transcribe
Listen to audio and write what you hear in text. Captures spoken language for training and research.
Record speech
Read a prompt aloud. The app checks audio quality automatically before submission.
Describe images
View a photo and speak or write what you see in your language. Grounds vocabulary in the real world.
Validate
Rate other contributors' work. Peer review is what makes the dataset trustworthy.
Languages
Every language matters, starting with the ones you speak.
7,000+ languages are supported worldwide. Native speakers contribute through translation, recording, and peer review, one language at a time.
7,000+
languages worldwide
5 task types
to contribute
100% open
research datasets
For organisations
Supply the data. Own the research.
Universities, NGOs, and research labs partner with All Voices to build datasets for their language community. Upload your source material, define the task types, and receive a reviewed, exportable dataset, without building contributor infrastructure yourself.
Apply to partnerApplications reviewed within 24 hours.
Project management
Create projects, define target languages, upload text, audio, or images as source material.
Contributor network
Tap into a growing community of native speakers across Africa.
Peer review pipeline
Submissions are validated by peers before entering your dataset.
Export on demand
Download approved datasets as JSON or CSV, with anonymised contributor demographics for research.
Role-based access
Invite your team as org admins. Super admins at Issaka AI handle platform governance.
The more you contribute, the further you climb.
Every approved contribution earns points โ translate, transcribe, record, describe, or validate. Keep a streak going and you earn bonus rewards on top. The top 100 contributors appear on the leaderboard each week, updated in real time so you can see exactly where you stand.
Leaderboard
This week| Rank | Contributor | Points |
|---|---|---|
| 1 | VCVictor Chukwuemeka | 312 pts |
| 2 | EAEvans Agyei | 201 pts |
| 3 | ADAmara Diallo | 98 pts |
Your rank: โJoin to see your rank
| Task | Points |
|---|---|
| Translate/Transcribe | 10 pts |
| Record speech | 15 pts |
| Image elicitation | 12 pts |
| Validate | 5 pts |
| Daily first contribution | +3 pts |
| 7-day streak | +25 pts |
Start contributing today.
Record, translate, and review in the languages you speak. Every contribution helps preserve African languages for research.