Mandarin AI Content Annotator (Singapore, C1 English)

About OpenTrain

OpenTrain is the #1 platform for building careers in AI training and data labeling. We connect skilled contributors with project work that helps shape how modern AI systems behave.

Joining OpenTrain means access to flexible, remote projects where your language skills and judgment directly improve real AI systems.

About AI training work

AI training (also called data labeling or human feedback work) is the human side of building artificial intelligence. Contributors review and rate examples—text, images, audio—that teach models to answer accurately, safely, and usefully.

This work is remote, often part-time, accessible to people with strong language skills and attention to detail, and places you on the cutting edge of how AI is developed.

The role

We are hiring Mandarin (Chinese) native-level annotators based in Singapore to evaluate generative AI outputs. You will work as an independent contractor on a project basis, focusing on text evaluation with occasional image or video review.

This is an entry-level professional role for contributors with at least one year of professional experience, a bachelor’s degree, and strong critical thinking and fact-checking abilities.

What you'll do

Follow detailed written guidelines to evaluate model outputs and apply labels consistently. Tasks are project-based and may vary over time, so flexibility and reliability matter.

• Rate output quality and factual accuracy for primarily text-based AI responses.
• Compare alternative responses and choose or rank the best ones.
• Classify and label content according to project taxonomies (classification, rating, RLHF, SFT-style prompts).
• Perform prompt-response writing and question-answering tasks when requested.
• Identify and flag safety, policy, or sensitive-content issues; participate in red-teaming tasks as needed.
• Occasional review of images or short videos alongside text, following the same accuracy and safety standards.

Requirements

You must meet all of the following to be considered for this role. We will verify qualifications during skills screening and calibration.

• Current residence in Singapore (required) and legally able to work as an independent contractor/freelancer.
• Native-level Mandarin Chinese (reading, writing, comprehension).
• English proficiency at C1 or higher.
• Bachelor’s degree (any field) and 1+ year of professional work experience.
• Strong critical thinking, attention to detail, and fact-checking abilities.
• Comfortable following complex written guidelines and asking clarifying questions.
• Able to review potentially sensitive or disturbing content when required.
• Reliable computer and stable high-speed internet.
• Willingness to pass language/skills screening, calibration, and ongoing quality checks.

Who should apply

This role is a good fit for professionals and freelancers who want flexible, remote, part-time work and who care about AI quality and safety. Ideal applicants are precise communicators with strong bilingual skills and experience working to standards or guidelines.

• Entry-level pros with 1+ year of experience and a bachelor’s degree.
• Native Mandarin speakers living in Singapore who also operate fluently in English (C1+).
• People seeking steady project-based work at 20+ hours/week.

How it works — schedule, pay, and next steps

This is contract, part-time work with tasks assigned per project. Expect to commit 20+ hours per week when work is available; exact hours vary by project. You will be paid hourly at USD 12 per hour.

Selection includes a language and skills screening, a calibration exercise to ensure alignment with project guidelines, and ongoing quality checks. If you meet the requirements, you can apply and complete the screening to be matched with suitable projects.

• Employment type: Contractor, Part-time.
• Pay: USD 12 per hour, paid per contractor agreement.
• Time requirement: 20+ hours/week (project-dependent).
• Data types: Primarily text; occasional images or video.
• Labeling activities include classification, rating, RLHF, SFT-style prompt/response work, red-teaming, and text generation evaluation.