Annotation
AI Data Annotation and Labelling
Label eligible client-provided text, images, or other data using documented machine-learning guidelines.
Workers label data the client provides, following written guidelines. Human review and quality assurance are included when the task requires them. Unlawful, exploitative, or unsafe tasks are not accepted.
Who this is for
Workers label text, images, or other material you provide, using a written guideline. The guideline should be specific enough that two careful people would usually make the same choice. A sample review can be included when you ask for it.
What to include when you ask for a quote
Describe the data type, the label set, the number of items, and any rule for items that do not fit. Say how disagreements should be recorded. Do not send material that is unlawful, exploitative, or unsafe to handle.
What this service does not do
Annotation applies your guide to your data. It does not train a model on Earn.net, and it does not include requests to evade a safety system or to label material the site has refused.
Examples of a brief
- Apply labels to text snippets using a written guide
- Mark image regions described in the instructions
- Send a sample for quality review before the full batch
What happens after you enquire
The enquiry form saves your business name, contact details, the selected service, and your description. An administrator reads that record and can add a quotation, the agreed scope, and a follow-up date. Nothing on this page is a price or a delivery promise.
Agreed work can later be published as a task. Workers do not see your enquiry. They see a task only when it is published and they are eligible to accept it. Unlawful, exploitative, or unsafe work is refused. For the steps in more detail, read how Earn.net works and how to prepare a task brief.
This page does not state a price or a delivery time. Use the enquiry form to request a quote.