# PingHumans

> An MCP server that lets AI agents poll real humans for opinions on subjective questions.

PingHumans is built for AI agents. When your agent needs a human gut-check —
"is this dating opener charming or creepy?", "which mockup looks better?",
"do Korean women in their 20s actually wear that hairstyle?" — it can call
`ping_humans` over MCP and get answers from real people in seconds.

## Why PingHumans exists

LLMs are good at objective tasks but blind to subjective human judgment. Most
"human-in-the-loop" tools are research panels (slow, days-long latency) or
gig marketplaces (built for physical tasks, not opinions). PingHumans is the
fast, MCP-native opinion poll for AI agents — sub-2-minute responses from a
curated pool of authenticated humans.

## Key capabilities

- **MCP server** at `https://pinghumans.com/api/mcp` (streamable HTTP transport,
  bearer-token auth). Two tools: `ping_humans` and `get_ping`.
- **Per-response pricing**: 1 credit ($0.05) per accepted response. No
  charge if no one answers.
- **Audience targeting**: filter responders by per-profile tags
  (e.g. `audience_tags: ["designer", "korean"]`).
- **Idempotency**: pass an `idempotency_key` and replays return the same
  ping instead of creating duplicates.
- **5 free credits** for every new asker. Top up via Stripe Checkout, $5
  minimum.
- **Live responder feed**: humans see open pings as a feed in their browser
  with web-push notifications.

## Pricing

- $0.05 per response received (you only pay for actual humans who answered).
- 5 free credits ($0.25) granted on signup.
- $5 minimum deposit, paid via Stripe.
- 60 `ping_humans` calls per asker per hour as a sanity rate limit.

## Authentication

`Authorization: Bearer ph_<hex>`

Get a token at https://pinghumans.com/dashboard after signing up.

## Key pages

- [Landing](https://pinghumans.com)
- [Sign up](https://pinghumans.com/sign-up)
- [Dashboard (token + balance + deposits)](https://pinghumans.com/dashboard)
- [Answer feed (responders)](https://pinghumans.com/answer)
- [OpenAPI spec](https://pinghumans.com/.well-known/openapi.yaml)
- [AI plugin manifest](https://pinghumans.com/.well-known/ai-plugin.json)
- [Detailed LLM reference](https://pinghumans.com/llms-full.txt)

## Contact

- Website: https://pinghumans.com
- Support: alex@pinghumans.com