Skip to content
For developers

A form backend you call like any other API.

Post JSON, form data or files to one endpoint, get field errors back you can show, and receive every checked submission on your server as a signed webhook. Everything in the dashboard is in the API too.

REST API200 OK
# Create a form
curl -X POST https://formpik.com/api/v1/forms \
  -H "Authorization: Bearer fp_live_..." \
  -d '{ "name": "Quote request" }'

# Read what came in
curl https://formpik.com/api/v1/submissions \
  -H "Authorization: Bearer fp_live_..."

What you get

The parts of a form backend you would rather not write.

  • 01

    One endpoint for every form

    POST /api/in/{key} takes JSON, form posts and multipart uploads: up to 10 files of 10 MB each and 25 MB per request.

  • 02

    Errors per field

    Field rules run on the server. A broken rule answers 422 with a message for each field, and the validate address checks answers without storing them.

  • 03

    Safe to retry

    Send an Idempotency-Key and a retry never stores the same submission twice. Repeats within 10 seconds are dropped too.

  • 04

    Signed webhooks

    A Send webhook step posts each submission to your server with an HMAC SHA-256 signature and a timestamp you can check.

  • 05

    The whole product as an API

    Create and publish forms, connect sources, build flows and destinations, work the inbox and, with an admin key, run the team and billing under /api/v1, at 120 requests a minute.

  • 06

    No reload on the page

    One script tag sends the form in the background, shows your messages and errors under each field, and fires events you can listen to.

In code

Three calls cover most of it.

Send a submission, read the inbox and check that a webhook came from Formpik. The docs have every option, status and error code.

JavaScript
const response = await fetch("https://formpik.com/api/in/YOUR_SOURCE_KEY", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Idempotency-Key": "order-1042"
  },
  body: JSON.stringify({ name: "Jane Cooper", email: "jane@example.com", topic: "sales" })
})

const result = await response.json()

if (response.status === 422) {
  // result.errors is { field_key: "message" }, ready to show under each input
}

Getting started

From an API key to your first webhook.

  1. 01

    Create an API key

    In the workspace menu, open API keys. Pick read only, full or admin access; the key is shown once.

  2. 02

    Create a form and a source

    In the dashboard or with POST /api/v1/forms and /api/v1/sources. The source answers with its endpoint address.

  3. 03

    Post submissions

    From the browser, your backend or a queue. Show the field errors a 422 answers with.

  4. 04

    Receive them

    Add a Send webhook step to a flow, check the signature on your server, and read anything back with the API.

Questions

What developers ask first.

Is there an SDK?
Not yet, and you do not need one: every endpoint is plain HTTPS with JSON, so any language and any HTTP client works.
Can I call it from the browser?
The submission endpoint, yes: it answers CORS requests from the websites you allow. Keep API keys on your server and never put them in a page.
How do I test without filling my inbox?
Send test submissions with the API or test one flow on sample answers. Both are simulated, deliver nothing and do not count toward your monthly submissions.
What are the limits?
120 API requests a minute per key, 120 posts a minute per address on the endpoint, and the limits of your plan for forms, flows and monthly submissions.
Can my AI agent use it?
Yes. Agents connect to the MCP server, which offers every endpoint as a tool. They sign in with a link, or use the same API keys.

The same product

Same forms and flows, another way in.

Whoever builds them, the forms, flows and inbox are the same. A team can mix all three.