Verbatim transcription
Write down exactly what speakers say, following the project's style rules.
Speech recognition models learn from audio paired with accurate transcripts. Transcribers listen to recordings and write exactly what was said, following the project's rules for punctuation, filler words, overlapping speech and noise. One6 matches transcribers to AI speech projects in their native languages.
Write down exactly what speakers say, following the project's style rules.
Mark who is speaking and when, and segment long recordings.
Check existing transcripts against the audio and correct errors.
Tag noise, music, laughter and other non-speech sounds.
Share your location, expertise, skills, experience, languages and work preferences.
Complete a skills test or trial task. When a project needs your profile, we contact you with the details.
Review the project's instructions and terms, including how work is reviewed and paid, and train on its guidelines.
Work remotely. Your accuracy and output are tracked, and contributors below the quality bar are retrained or removed.
The core skill is the same, but AI projects usually have strict rules for how to mark filler words, false starts, overlapping speech and background noise, because the transcript is training data rather than a readable document.
Yes. One6 contributors work remotely, from wherever they are.
Compensation depends on the project. Before you start any project you receive its instructions and terms, including the tasks, how work is reviewed and how it is paid, and you are free to decline.
No. Contributors are selected for each project based on its requirements, such as languages, domain and experience. We contact you when a project matches your profile.
Read the Contributor Terms and Privacy Policy, learn more about One6, or see all remote AI training jobs.
Share your information once and we'll contact you when a project matches your profile.