Try the API before you wire it in.

DemoAPI gives project APIs a public page with readable docs, a controlled request console, and code you can copy.

Start with a public HTTPS endpoint you control. Listings are reviewed by a person before they go live.

Built-in example / Local model health

live request / safe fixture same origin
GET/api/demo/model-health
Response

200 fixture preview

{
  "status": "ready",
  "model": "demo-7b-instruct",
  "runtime": "local",
  "queue_depth": 0,
  "checked_at": "2026-07-15T14:32:00.000Z"
}
  • Sample data is labeled
  • Runs only when you press the button
  • No account needed

APIs you can run now

Each reference demo uses a fixed, first-party handler. Nothing here can reach your server, machine, or private network.

01
GET

Local inference

Local model health

/api/demo/model-health

Check whether a small local model service is ready to take work.

Open
02
POST

Computer vision

Vision job preview

/api/demo/vision-job

Preview how an image inspection job would be accepted and scheduled.

Open
03
POST
/api/demo/robot-route

Validate a short waypoint plan before handing it to a robot controller.

Open

Give your API a front door.

Send us the public docs and one safe route. We check the claim, agree on limits, and build the demo page.

  1. 01 / SUBMITShare the public endpoint and what it does.
  2. 02 / REVIEWWe confirm ownership, scope, and safe example data.
  3. 03 / LIMITMethods, fields, payload size, and request caps stay narrow.
  4. 04 / PUBLISHYou get one public page with docs and a runnable console.
Request a listing

Public enough to try. Narrow enough to control.

A publisher chooses the exact public route, method, fields, and example values. DemoAPI does not offer shell access, filesystem access, or an open proxy.

Endpoints
Verified public HTTPS only
Routes
An explicit allowlist per project
Secrets
No production credentials in a public demo
Limits
Small payloads and request caps
Read the trust notes

One small site in a larger workshop.

DemoAPI.org is built by AI Embedded Systems. It sits alongside AI-WF, our workflow studio, as part of the same public project catalog.