VOICE SFT
Setting things up for you...
Choose your Voice SFT path

Your voice has
value.

Find eligible voice and language projects, complete guided tasks, and help improve speech technology with your natural voice.

For clients Build a voice data project Request managed collection for speech, language, and contributor workflows. Start a client request → For workers Start or continue tasking Access your assigned projects, recordings, and contributor dashboard. Go to worker sign in →
SMKJ
Join 400+ contributors worldwide
Sound like you. That’s enough.
Contributor experience
A workspace that tells you what to do next.

Voice SFT is designed to keep contributors oriented across project availability, task requirements, review status, and payment guidance.

Clear project statesAvailable, locked, in review, completed, and out-of-task states are shown in plain language.
Focused task screensEach project step keeps instructions close to the action so you can complete work without hunting around.
Support when neededIf something blocks your progress, support stays part of the workspace instead of sending you to a dead end.

How tasking works

Every project has its own requirements. Your dashboard shows what is available, what needs an eligibility step, and what happens after you submit.

Choose an available project

Open your dashboard to see active projects, estimated time, status, and any requirements. If a project is locked, we explain what is needed first.

Complete the guided steps

Follow the project instructions one step at a time. Depending on the project, you may record audio, complete verification, answer prompts, or submit a short video.

Track review and payment

After submission, your project status updates to show review, completion, or out-of-task states. Approved work is paid through the platform where you were hired.

Short Scripts
Read short prompts aloud — quick, simple, done in minutes.
Narrations
Speak naturally for longer takes — stories, guides, and more.
Conversations
Record real back-and-forth dialogue with another contributor.
/æ/ /iː/ /ɔː/ /ʌ/
Phonetics
Pronounce specific sounds to help train accent-aware AI models.

Diverse projects for real voices

Voice SFT projects are built around natural speech, local languages, and real-world accents. When a project matches your profile, your dashboard shows what is available and what to do next.

  • No professional equipment needed
  • Guided tasks with saved progress
  • Payment handled through your hiring platform
View Worker Dashboard

Voices we're looking for

We collect recordings across a growing range of languages and accents. If you speak any of the following, there's a project waiting for you.

🇸🇦 Arabic 🇬🇧 English 🇫🇷 French 🇩🇪 German 🇪🇸 Spanish 🇮🇹 Italian 🇵🇹 Portuguese 🇳🇱 Dutch 🇹🇷 Turkish 🇵🇱 Polish 🇯🇵 Japanese 🇰🇷 Korean 🇦🇪 Emirati Arabic 🌍 Gulf Arabic 📖 Modern Standard Arabic + More coming soon 🇸🇦 Arabic 🇬🇧 English 🇫🇷 French 🇩🇪 German 🇪🇸 Spanish 🇮🇹 Italian 🇵🇹 Portuguese 🇳🇱 Dutch 🇹🇷 Turkish 🇵🇱 Polish 🇯🇵 Japanese 🇰🇷 Korean 🇦🇪 Emirati Arabic 🌍 Gulf Arabic 📖 Modern Standard Arabic + More coming soon
Contributor fit

Natural speech is the point.

Projects are matched by language, region, device, and eligibility needs. When a task appears in your dashboard, use your normal speaking style and follow the project instructions closely.

01
Use your real accentSpeak naturally instead of performing. Local pronunciation and everyday pacing are useful for voice data.
02
Record with clarityChoose a quiet moment, check the script, and use the device guidance shown in the project.
03
Follow project requirementsSome projects may require a specific language, country, verification step, or recording format.

Types of voice work

From a quick two-sentence read to a full conversation session — there's a task for every schedule and comfort level.

Recording
Voice Acting

Read scripted lines with natural expression — characters, narrations, product announcements, and more.

Scripted
Conversation
Simulated Voice Calls

Take part in structured two-way conversations to train conversational and call-centre AI systems.

Paired
Text output
Audio Transcription

Listen to short audio clips and type exactly what you hear — improving speech-to-text across accents.

Listening
Annotation
Audio Labeling

Tag audio by tone, emotion, noise level, or speaker count. Quick tasks, completable in short bursts.

Quick tasks
Pronunciation
Phonetic Recordings

Pronounce specific words or sounds to help train pronunciation models. Accents especially valued.

Accent
Beginner
Short Script Reading

Read one or two sentences at a time — headlines, everyday phrases. The easiest way to start.

Solo
Recording
Voice Acting

Read scripted lines with natural expression — characters, narrations, product announcements, and more.

Scripted
Conversation
Simulated Voice Calls

Take part in structured two-way conversations to train conversational and call-centre AI systems.

Paired
Text output
Audio Transcription

Listen to short audio clips and type exactly what you hear — improving speech-to-text across accents.

Listening
Annotation
Audio Labeling

Tag audio by tone, emotion, noise level, or speaker count. Quick tasks, completable in short bursts.

Quick tasks
Pronunciation
Phonetic Recordings

Pronounce specific words or sounds to help train pronunciation models. Accents especially valued.

Accent
Beginner
Short Script Reading

Read one or two sentences at a time — headlines, everyday phrases. The easiest way to start.

Solo
Annotation
Audio Labeling

Tag audio by tone, emotion, noise level, or speaker count. Quick tasks, completable in short bursts.

Quick tasks
Pronunciation
Phonetic Recordings

Pronounce specific words or sounds to help train pronunciation models. Accents especially valued.

Accent
Beginner
Short Script Reading

Read one or two sentences at a time — headlines, everyday phrases. The easiest way to start.

Solo
Recording
Voice Acting

Read scripted lines with natural expression — characters, narrations, product announcements, and more.

Scripted
Conversation
Simulated Voice Calls

Take part in structured two-way conversations to train conversational and call-centre AI systems.

Paired
Text output
Audio Transcription

Listen to short audio clips and type exactly what you hear — improving speech-to-text across accents.

Listening
Annotation
Audio Labeling

Tag audio by tone, emotion, noise level, or speaker count. Quick tasks, completable in short bursts.

Quick tasks
Pronunciation
Phonetic Recordings

Pronounce specific words or sounds to help train pronunciation models. Accents especially valued.

Accent
Beginner
Short Script Reading

Read one or two sentences at a time — headlines, everyday phrases. The easiest way to start.

Solo
Recording
Voice Acting

Read scripted lines with natural expression — characters, narrations, product announcements, and more.

Scripted
Conversation
Simulated Voice Calls

Take part in structured two-way conversations to train conversational and call-centre AI systems.

Paired
Text output
Audio Transcription

Listen to short audio clips and type exactly what you hear — improving speech-to-text across accents.

Listening
For enterprise customers

Build speech datasets with a managed contributor network.

Voice SFT runs managed multilingual voice and language data collection, handling contributor operations, guided workflows, review, and delivery for AI teams.

Explore customer solutions Request a project
Language coverageActive sourcing
Quality reviewHuman checks
Contributor operationsGuided steps
Talent experience

A clearer way to find, complete, and track voice work.

Voice SFT gives contributors a focused workspace for eligible projects, guided task steps, submission status, and payment expectations, so every project feels easier to understand from the first click.

Contributor portal

Simple paths, fewer dead ends, and clearer next steps.

Whether you are starting a recording task or returning to check progress, the dashboard is designed to show what is available, what is locked, what is under review, and what happens next.

Go to worker sign in
01

Eligible projects

Available projects are shown separately, with locked projects clearly marked when another step must be completed first.

02

Guided tasking

Each project opens into a focused flow with only the instructions, actions, and controls needed for that step.

03

Status clarity

Submissions, reviews, completed work, and out-of-task states are labeled clearly so contributors know where they stand.

04

Payment expectations

After approval, payment guidance reflects the platform where the contributor was hired, keeping the final step straightforward.

Ready to continue
your Voice SFT work?

Open your contributor portal to view available projects, continue in-progress work, and check your submission status.

Open Worker Portal
Works from your phone or laptop - no special equipment required