What the work involves

You record from supplied scripts — conversational turns, instructional copy, narrative passages, and short CX-style utterances — and deliver clean audio files that meet a fixed technical spec. Sessions are self-directed: you receive scripts and guidelines, record in your own space, check for mouth noise, plosives, room reflections and level drift, then upload. Expect repeated takes of the same line with deliberate variation in emotion, emphasis, and pace, and expect consistency notes across weeks, since a cloned voice model is unforgiving about session-to-session shifts in mic distance, energy, or accent drift.

The defining condition of this quest is voice cloning. The client will train a synthetic voice on your recordings for their internal customer-experience agent. The listing states this is exclusive to that use case and not sold, licensed, or reused elsewhere — read the consent terms yourself before applying, and skip this one if cloning is a line you don't cross.

What the platform screens for

  • Accent verification. Mercor screens for a genuinely neutral, standard UK English accent with UK residency. Strong regional colouring, or a UK accent acquired non-natively, tends to be filtered early.
  • Recorded evidence. A reel or samples from voice acting, dubbing, narration, podcasting, or broadcast work, plus specifics about your chain: microphone, interface, treatment, and typical noise floor.
  • Direction-taking. Screens for TTS work usually probe whether you can hit a precise read repeatedly rather than improvise — precision under repetition beats performance flair here.
  • Availability honesty. 5–10 hours per week, sustained, matters more than a large one-off burst.

Logistics

Fully remote and asynchronous, recorded from your own space on your own schedule within submission deadlines. Observed pay for this listing sits in the $50–100/hr band — as posted, not guaranteed, and often tied to accepted audio rather than raw hours. Project duration is defined by the client's dataset needs; treat it as a multi-week engagement rather than ongoing work, and expect QA rejections on files that miss the noise or consistency spec.