Founding Trainer list open

The human data network for physical AI.

First-person demonstrations of everyday tasks, recorded in real homes. Robotics teams get training-ready data. Trainers earn USDC with nothing but a phone.

Launch rates
Up to $34/hr
Paid in
USDC
Payouts
Within 3 days
KitchenZ/02 · Z/03
Z/02
Rinsing a glass
Z/03
Slicing a tomato
Z/02
Rinsing a bowl
Z/03
Dicing cucumber
Z/03
Chopping greens
Z/03
Cutting tomato
LaundryZ/01
Z/01
Folding a shirt
Z/01
Filling a pillowcase
Z/01
Smoothing a sheet
Z/01
Folding cloth
Z/01
Sorting a basket
Z/01
Unloading a washer
TidyingZ/04 · Z/06
Z/06
Sorting small parts
Z/06
Organising a drawer
Z/04
Spray and wipe
Z/06
Filling a basket
Z/04
Wiping a table
Z/06
Arranging a drawer

Example footage from stock libraries, shown until launch.

AI learned to write from the internet. Robots learn to work from people.

$6Binvested in humanoid robots in 2025
+50%task success NVIDIA reported from 20,000+ hours of first-person video
$15–40paid per hour of first-person footage by robotics teams

Sources: MIT Technology Review, Washington Post and robotics data vendors, 2026.

The data

Real homes. Real hands. Real variety.

Point-of-view footage of the tasks robots struggle with most, from kitchens, bedrooms and living rooms around the world. Example clips shown until launch.

1080p
30 fps video
42
hand keypoints a frame
2×
checks on every clip

Washing up and food prep, at counter height, from the trainer's own eyes.

Slicing a tomato
Dicing cucumber
Rinsing a glass
Chopping greens
Rinsing a bowl
Cutting tomato

Folding and handling fabric. Deformable objects are among the hardest things for robots to learn.

Folding a shirt
Filling a pillowcase
Smoothing a sheet
Folding cloth
Sorting a basket
Unloading a washer

Sorting, organising and wiping down. The precise, everyday movements home robots need most.

Sorting small parts
Organising a drawer
Filling a basket
Arranging a drawer
Spray and wipe
Wiping a table

Need a task that isn't covered? Describe it and we open a mission for it.

Example request

Restocking shelves in small retail stores. 200 hours, Southeast Asia, both hands visible.

Request a mission

Example footage from stock libraries, shown until launch.

How it works

Record it. Review it. Get paid.

  1. 01

    Pick a mission

    Every mission comes with a brief, an example clip and a rate.

  2. 02

    Record on your phone

    Mount your phone, do the task, upload. Checks run before anything is sent.

  3. 03

    Pass review

    Your phone checks the clip first, then a person reviews it to confirm it’s clear, private and yours.

  4. 04

    Get paid in USDC

    Approved clips are paid on Solana, usually within 3 days of approval.

The app

First clip to first payout, in minutes.

The real Zeth trainer app. Choose a mission, upload your clip, track review and watch USDC arrive.

zethnetwork.com/app

The real Zeth trainer app, recorded in demo mode

Launch missions

Everyday tasks. Paid by the clip.

Launch rates up to $34/hr in the US. Adjusted by region, shown before you record.

  • Fold laundryShirts, towels or trousers, one at a time8–12 min≈ $27/hr$4.50 per clip
  • Load the dishwasherPlates, cups and cutlery with the rack in view4–6 min≈ $27/hr$2.25 per clip
  • Prepare a simple mealChopping, spreading or assembling on a counter8–12 min≈ $27/hr$4.50 per clip
  • Clear and wipe a tableItems moved off, then a full wipe-down3–5 min≈ $28/hr$1.88 per clip
  • Open and close containersJars, bottles, zips and lids, five minimum2–4 min≈ $30/hr$1.50 per clip
  • Sort small objectsMixed items sorted into groups3–5 min≈ $28/hr$1.88 per clip

Launch rates for US trainers, paid per accepted clip. Hourly figures are based on average clip length. Rates are adjusted for your region and shown in the app before you record.

For robotics teams

Tell us the task. We'll film it.

Training-ready, first-person episodes from homes around the world. Labelled for your pipeline and collected against your brief.

episode.jsonSample schema
{
  "episode_id": "zeth_ep_000001",
  "mission": "Z/01 fold_laundry",
  "video": { "res": "1920x1080", "fps": 30, "view": "head" },
  "context": {
    "country": "KE",
    "setting": "home/bedroom",
    "objects": ["t-shirt", "towel"]
  },
  "steps": [
    { "t": [0.0, 4.2], "text": "pick up shirt by shoulders" },
    { "t": [4.2, 9.8], "text": "fold sleeves to center" }
  ],
  "hand_pose": { "format": "21kp_xyz", "hands": 2 },
  "qa": { "auto": "pass", "human": "pass" }
}

Illustrative format. The final schema is shaped with launch partners.

Trust

Private by design.

Privacy

No faces. No audio.

Hands and the task only. Audio is removed, and faces and screens will be blurred before any data is shared.

Control

You approve every clip

Preview before you submit. Delete any time until it's approved.

Payment

Paid in dollars

Every payout is USDC. No token required to earn.

Proof

Verifiable onchain

Payouts settle on Solana, so anyone can check that trainers get paid.

Roadmap

Where Zeth goes next.

  1. Phase 1 · Now

    Collect

    Founding Trainers, first missions, fast USDC payouts.

  2. Phase 2

    Prove

    An open sample dataset anyone can benchmark.

  3. Phase 3

    Demo

    An open robot policy trained on Zeth footage.

  4. Phase 4

    Sell

    First robotics customers. Revenue flows to trainers.

  5. Phase 5

    Grow

    Workshops, warehouses and farms, worldwide.

Questions

Questions, answered.

What do I need to start?

A smartphone that records 1080p video and a way to mount it on your head or chest. A basic head strap costs around $15.

How much can I earn?

During launch, missions pay up to $34/hr in the US, based on average clip length. Each mission shows exactly what it pays before you start, adjusted for your region. Launch rates are boosted and will come down as Zeth grows.

Who uses the footage?

Robotics teams use it to train robots to do everyday physical tasks, learning from how real people move their hands in real homes around the world.

Why would a clip be rejected?

Hands out of frame, poor lighting, a face or personal details in shot, footage that was already submitted, or footage that isn't yours.

Is there a token?

Not yet. Zeth plans a community token, and trainers are paid in USDC either way, so you never need to buy or hold it. The only official contract address will be posted on this site and @zethnetwork.

When does it launch?

Founding Trainers get the first missions before anyone else. Join the list and we'll email you when they open.

Get started

The robots are coming. This time, the people who teach them get paid.

The first 1,000 Founding Trainers get early access to missions, priority review and a permanent badge.

I want to

Launch and mission updates only. Unsubscribe any time. See our privacy policy.