Luel (Luel.ai) is a marketplace designed to collect and license high-quality, rights-cleared, multimodal training data for artificial intelligence models (like computer vision, embodied AI, and large language models).
Earn Money Speaking isiZulu, isiXhosa, Afrikaans and English Online
What Is Luel AI?
Luel AI is an AI data platform that pays contributors to complete online tasks such as conversational recordings, audio scripts, and document collection projects. The platform works with multilingual contributors from around the world to improve voice recognition and language technology.
For South Africans, some of the most exciting opportunities currently available include isiZulu and isiXhosa conversational recording tasks that pay in US dollars.
Available Online Audio Tasks
isiZulu Conversational Audio Task
Participants can earn:
- $15 per hour (around R270 per hour depending on exchange rates)
- $0.25 per minute recorded
The task involves having a natural conversation in isiZulu with another person using Luel AI’s online recording platform.
Topics can include:
- Daily life
- Work
- Hobbies
- Opinions
- General conversations
Both participants receive payment for the completed recording session.
isiXhosa Conversational Audio Task
The isiXhosa task works the same way and also pays:
- $15 per hour
- $0.25 per minute
Contributors must speak natural isiXhosa during the conversation while keeping background noise to a minimum.
English Conversational Audio Task
South Africans who are fluent in English can also participate in English conversation tasks.
The English project currently pays:
- Around $20 per hour
- Approximately $0.33 per minute
All English accents are accepted, including South African English accents.
How the Recording Process Works
The process is simple and fully online.
Step 1: Create an account on the Luel AI contributor platform.
Step 2: Choose an available task.
Step 3: Invite another person to join the recording session.
Step 4: Both participants connect through the online platform.
Step 5: Record a natural conversation lasting between 2 minutes and 1 hour.
Step 6: Submit the recording and wait for approval and payment.
The platform encourages authentic conversations instead of scripted recordings.
Requirements to Participate
To qualify for these online opportunities, applicants generally need:
- Internet access
- A smartphone, laptop, or computer
- Headphones
- A quiet recording environment
- Fluency in the required language
- A conversation partner for the audio tasks
For isiZulu and isiXhosa projects, contributors should preferably be native or fluent speakers.
Important Recording Tips
Luel AI recommends the following for successful submissions:
- Wear headphones to reduce echo
- Record in separate rooms if possible
- Avoid TV or music in the background
- Speak naturally
- Do not read from scripts
- Keep the conversation authentic
Good-quality audio improves the chances of approval and payment.
Why This Opportunity Matters
Remote online jobs and side hustles are becoming increasingly popular in South Africa, especially among unemployed youth, students, and people looking for extra income.
What makes this opportunity attractive is that:
- It is fully online
- You can work from home
- No formal qualifications are required
- Multiple South African languages are accepted
- Payments are made in dollars
- Flexible participation
This is especially useful for people who are comfortable speaking isiZulu, isiXhosa, or English and want to use their language skills to earn money online.
Final Thoughts
AI companies are investing heavily in African languages and accents, creating new online income opportunities for South Africans. Through platforms like Luel AI, ordinary conversations can now become a source of income.
If you have a stable internet connection, a quiet place to record, and someone to partner with, this could be a simple way to earn extra money remotely while contributing to the future of AI technology.
Because the “easy” data available publicly on the internet is largely exhausted or wrapped in copyright issues, Luel acts as the bridge between AI development teams who need niche data and everyday contributors who get paid to provide it.
It operates on a dual-sided model. Here is exactly how it works for both sides of the marketplace:
1. For Contributors (The Data Creators)
Anyone with a smartphone or computer can sign up to become a contributor. Instead of passive data scraping, Luel relies on explicit, consensual data creation.
-
Task Assignment: Once registered, contributors browse available tasks on their phones. These tasks vary widely based on what AI companies are currently buying. Common jobs include recording real-life daily conversations, photographing specific objects, or filming mundane daily activities (like cooking, walking, or using tools).
-
Continuous “Computer-Use” Capture: For building software-automation AI, some tasks involve running background tools that securely log screen recordings, user keystrokes, and workflow boundaries while a professional works in specialized desktop apps.
-
Sensor Pairing: For robotic and embodied AI training, the mobile app often pairs first-person video capture with raw IMU (Inertial Measurement Unit) sensor data from the phone to track physical motion and orientation.
-
Payouts: Contributors submit their files through the platform. Once the data passes quality checks and is verified as “clean” and right-cleared, the user is paid via platforms like PayPal or Wise.
2. For AI Developers & Enterprises (The Data Buyers)
AI labs, tech companies, and researchers use Luel to source the massive amounts of niche info required to make their models smart.
-
The Curated Catalog: Buyers can browse a marketplace of pre-collected, pre-labeled, and completely compliant datasets.
-
Custom Campaigns: If a team is training an AI for a specific medical, industrial, or language niche, they can commission a custom campaign. Luel then pushes those specific requirements out to its global network of contributors (spanning over 90+ countries) to crowd-source the data.
-
Rights Clearance: The most critical legal aspect of Luel is that every piece of data is “rights-cleared.” Because contributors explicitly sign off on how their data will be used, AI companies can train commercial models on it without fearing copyright lawsuits or privacy violations.
How the Data is Structured
Luel uses natural language processing (NLP) and automated pipelines to instantly structure incoming files. For example, if a contributor uploads a raw video of a mechanic fixing a car, Luel’s background system automatically tags the workflow boundaries, labels the specific tools being used, and matches the audio text to the exact seconds of the video—turning raw media into AI-ready, structured training material.
