AI Voice Data Jobs: Recording, Speech and Audio Work

AI voice jobs by type of work

Speech models learn from people talking, and voice products are judged by people listening. The table groups the open listings we read on 14 November 2026 by the kind of work. Programs are alphabetical within each row; pay is as each listing published it.

Sources, read 14 November 2026: Alignerr Voice Recording Specialist, Alignerr UK voice actor, Alignerr Italian voice actor, Alignerr Indian English voice actor, Alignerr Korean voice evaluator, Alignerr Saudi Arabic voice trainer, Alignerr Audio Transcription Specialist, Alignerr Hindi transcription; Handshake AI vocal arts, Handshake AI music analyst, Handshake AI audio engineers; Mercor UK voice cloning, Mercor Swiss German voice cloning, Mercor audiobook QA (German), Mercor audiobook QA (US), Mercor TTS French, Mercor TTS Brazilian Portuguese, Mercor lyrics (Thai), Mercor lyrics (Japanese), Mercor music producer; micro1 Hindi, micro1 English voice, micro1 music annotation; xAI Spanish tutor.

AI voice trainer and voice recording jobs

Home recording roles are the easiest to start. Alignerr's Voice Recording Specialist listing says "No voice acting experience required" and asks for "A microphone, headset, or smartphone capable of capturing good audio quality" (Alignerr). Handshake AI's vocal arts role says "Must have audio recording hardware and software." (Handshake AI).

Some recording work is a fixed job rather than hourly. Welo Data's French (Canada) project pays a "Fixed rate of $32" for "Approx. 50 minutes of total recorded audio" (Welo Data). Phone-based sentence recording can pay per unit: an Appen posting for Mexican Spanish paid "$0.030 USD per completed unit" (Appen). Our page on getting paid to record your voice for AI covers those app-based programs, voice licences and consent rules; this page does not repeat them.

Studio voice acting and voice cloning

Professional voice actors see the highest published rates. Alignerr's UK English role pays $120-$200 an hour for "in-person studio recording sessions at a professionally equipped facility local to you", in "2–3 sessions of 4 hours each", and states "Important: This project does not involve voice cloning." (Alignerr).

Mercor's CX Agent roles are cloning projects, for an AI customer-service voice. Its UK listing pays $50-$100 an hour, says "The engagement itself is a single recording session of approximately 4 hours", and that "Your voice recordings will be used exclusively for the client's internal CX AI Agent." (Mercor). A single session at a high rate is still a single session. Before you record, read the licence; the voice recording guide explains what to look for.

Speech annotation, transcription and audiobook QA

Transcription for AI is harder than office dictation. Alignerr's Hindi role describes audio "featuring multiple speakers, crosstalk, background noise, filled pauses" and asks for "Commitment : 10–40 hours/week" (Alignerr). Mercor's audiobook reviewers listen to AI narration and log "skipped, added, or mispronounced words" at "roughly 20 hours per week" (Mercor).

Language voice evaluation is close to coaching: micro1's Hindi role asks you to "Listen to audio clips in Hindi and evaluate the nativeness and fluency of the speaker or model output" (micro1), and xAI's language tutors "Help produce clear, natural audio data that accurately captures linguistic and prosodic details such as intonation, rhythm, and accent" (xAI). For every language and its rates, see bilingual AI training jobs.

Jobs for musicians and audio engineers

Musicians, producers and sound engineers have their own listings. micro1's Music & Sound Annotation role asks you to "Listen to 30–90 second field recordings" and label what you hear, and requires "3+ years of hands-on music or audio experience" (micro1). Mercor's Music & Lyrics roles pay per task, and "your per-task rate is based on your location" (Mercor). Handshake AI's Audio Engineers listing reaches up to $125 an hour for US-authorized engineers (Handshake AI).

Onsite speech jobs

Not all of it is remote. Welo Data's Speech & Voice AI Analyst roles list "Pay Rate: $28/hour", "Employment Type: W2 Full-Time Employee" and state "This is a 100% onsite position", in named US cities (Welo Data). That is employment, not contract work. For recording video of tasks rather than audio, see physical AI recording work.

If you have a trained voice, a rare accent or audio production experience, tell us about it and we'll email you the programs open to you.

Next step

Check pay and payout rules before you record: AI trainer pay rates and how AI training programs pay.

Tell us what you do. We'll send you the paid AI work that fits your field, wherever you live. It's free.

Origin Data Partners is independent of the programs on this page. A program may pay Origin when Origin introduces someone who joins it, which never changes what you are paid (how Origin is paid).