How PierNode Works

PierNode is an API-first platform for AI media generation. You integrate against one REST API, and PierNode handles authentication, metering, billing and the asynchronous work of running generations.

How a request flows

1. You create an API key in the PierNode console.
2. Your application calls the API with that key.
3. PierNode prices the request and reserves the cost from your balance.
4. The generation runs asynchronously.
5. You poll the task until it finishes, then download the result.

Results are collected by polling. There is no webhook callback for generations; if you need one, poll from a background job and notify yourself.

What is not available

Being explicit about the edges, so you do not design around something that is not there:

• No webhooks. Poll the task.
• No streaming or partial results. A task is pending until it is finished.
• No batch endpoint. One generation per request.
• No cancellation once a task is processing. See cancelling.
• One API key per account. Rotating replaces it immediately.
• No browser access. The API is server-to-server; see authentication.

Versioning

Every endpoint lives under /api/v1. Within v1 we will add endpoints, add optional request fields, and add fields to responses — so parse responses leniently and ignore what you do not recognise.

What we will not do inside v1 is remove a field, change what one means, or make an optional field required. A change like that would arrive as a new version alongside this one, with notice, rather than silently. We are not committing to a support window for v1 yet; we are committing not to break it underneath you.

Providers

Which model runs a generation is PierNode's concern, not yours. The API exposes no provider, model or checkpoint selection, and none appears in a response. That is deliberate: it keeps your integration stable while the machinery behind it changes.