Docs

API docs

Create a prediction, get the result, integrate it.

Request API Key

Call models with a FounderHooks key

Users only receive our key. The server forwards requests with the Replicate token configured by FounderHooks.

Proxy layer

A simple Replicate API wrapper

Users call the FounderHooks API, not Replicate directly.

The server forwards with REPLICATE_API_TOKEN.

Each request logs key, model, prediction id, and upstream status.

Predictions lifecycle

FounderHooks accepts your API key, forwards the request to Replicate, and returns the prediction payload.

Create a prediction with model + input payload.

Poll status through the FounderHooks prediction id endpoint.

Your Replicate token stays on the FounderHooks server.

Streaming output

Language and voice models often benefit from streaming. The platform UX should treat streaming as a first-class product surface, not a hidden SDK detail.

Stream partial tokens or intermediate events.

Surface partial output in the playground immediately.

Mirror the same behavior in the API snippets.

Webhooks and async delivery

For video, image batches, or multi-step runs, webhook delivery becomes the default integration surface. FounderHooks keeps the webhook pattern obvious in both docs and example code.

Attach a callback URL on prediction create.

Retry delivery on transient failures.

Log event receipts per prediction for debugging.

curl

API sample

curl -s -X POST https://www.founderhooks.com/api/v1/predictions \
  -H "Authorization: Bearer $FOUNDERHOOKS_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "black-forest-labs/flux-schnell",
    "input": {
      "prompt": "Generate a polished product visual for an AI platform homepage"
    },
    "wait": 10
  }'

JavaScript

API sample

const response = await fetch("https://www.founderhooks.com/api/v1/predictions", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${process.env.FOUNDERHOOKS_API_TOKEN}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    model: "black-forest-labs/flux-schnell",
    input: {
      "prompt": "Generate a polished product visual for an AI platform homepage"
    },
    wait: 10,
  }),
});

const prediction = await response.json();