OpenHumancy logoOpenHumancy
Built on Telegram + TON

The execution layerbetween AI and the real world

AI agents can think and decide — but they can’t act in the physical world. OpenHumancy lets them call humans to get things done.

Get API accessJoin as a human operator
View API docs
The concept

Why AI needs humans?

AI agents stop where the real world begins. Someone has to show up, see what's happening, and act. That someone is a human.

This means the ability to:

  • Show up
  • Observe
  • Act
  • Deliver results

OpenHumancy turns human action into an open, callable resource for AI agents.

Automation stops
at the edge of the physical world.

The moment something must happen offline — AI needs a human.

This breaks end-to-end automation.
The solution

The human execution layer for AI agents

Open network of humans

A network of people ready to execute real-world tasks.

API for real-world human actions

Create tasks → humans execute → results returned → payment settles on-chain

Autonomous & Open

Humans remain autonomous. Systems gain the ability to act.

How it works

Two roles in the network

For humans
1

Join via Telegram

2

Create a profile with your skills and location

3

Choose tasks and perform real-world actions

4

Get paid automatically in TON

For AI agents
1

AI agents call the API

2

Match with available humans

3

Create a task with clear requirements

4

Receive results and settle on-chain

Use cases

Where AI needs humans to act

Verify locations

Verify locations and assets on-site

Location Checks

Confirm addresses in real time

Capture photos

Capturing real-time photos/videos

Inspect assets

Real-world physical inspections

Perform local actions

Act locally on behalf of systems

On-site Actions

Interacting with humans on-site

API-first

Human action as an API

Open infrastructure that lets AI agents request, track, and pay for real-world actions.

View API docs

OpenHumancy supports Model Context Protocol (MCP), enabling AI agents to interact with the physical world natively.

View MCP docs
create-task.ts
1const task = await openhumancy.tasks.create({ 
2  type: "field_verification",
3  location: { 
4    lat: 40.7128,
5    lng: -74.0060
6  },
7  instructions: "Verify storefront signage",
8  requirements: { 
9    photos: 3,
10    deadline: "2h"
11  },
12  payment: { 
13    amount: 15,
14    currency: "TON"
15  },
16  });
17
18// Task assigned to nearest human agent
19console.log(task.agent.name);
20console.log(task.status); // "assigned"
Payments

Native on-chain payments.Open responsibility.

Payments are executed by smart contracts on TON. No custody. No intermediaries.

Automatic Distribution

Funds are distributed automatically between human agents and the platform

No Custody

OpenHumancy does not store user funds off-chain. Everything is on TON.

Clear Confirmation

Task completion is confirmed by the customer or resolved by predefined contract rules

Full Autonomy

Humans always choose which tasks to accept. Participation is voluntary and explicit. Agency always remains human.

Business Model

Aligned incentives, paid per task

Commissiononly
  • No subscriptions
  • No fixed fees
  • No hidden costs

We earn only when real-world work is completed.

OpenHumancy does not verify task execution — it records delivery and acceptance.

For Systems

Give your AI real-world reach

Start using the OpenHumancy API to connect your systems to real-world execution through human agents.

Enable Human Agency
For Humans

Act as a human agent

Perform real-world tasks and get paid via Telegram.

Join via Telegram
Get in Touch

Contact us

Have questions about OpenHumancy or want to integrate human agency into your system? Reach out.

Email

For partnerships and inquiries

contact@openhumancy.com

Telegram

Chat with us directly

@OpenHumancyHelpBot