Developers

The creative studio behind one API call

Composite creative endpoints nobody else offers as a single call, in production since 2024. One generation model and one prepaid balance, across a catalog that keeps growing.

Get an API key Quickstart

One endpoint runs every feature

feature picks what to make and args carries that feature’s inputs. Adding a new feature to the platform adds a value for the first and a shape for the second; it never adds an endpoint you have to integrate against. Here is a logo reveal, start to finish.

1Submit the generation

Nothing is rendered while you wait, so this returns in milliseconds with an id. The balance is charged here. Reference files like the logo go up first through /v3/files/upload and come back as the file id you see in args.image.

curl -X POST "https://api.neurall.io/v3/generations" \
  -H "Authorization: nl-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "feature": "video/logo-reveal",
    "args": {
      "image": "f_9c21e0b4",
      "prompt": "A blue anamorphic flare streaks across the black frame; the logo materializes in its wake.",
      "duration": 4
    }
  }'

{ "generation": { "id": "9f2c1d84…", "status": "pending", "balanceUsed": 0.64 } }

2Collect it when it lands

Poll the id every few seconds, 5 to 10 for a minute-scale render like this one, or pass a webhookUrl on the submit and let the finished generation come to you signed. A reveal takes two to three minutes.

# repeat every 5s or so until status is done, error or timeout
curl "https://api.neurall.io/v3/generations/9f2c1d84-6a3b-4f21-9c77-2f0a1b8e5d43" \
  -H "Authorization: nl-YOUR_API_KEY"

{
  "generation": {
    "status": "done",
    "output": { "video": "https://cdn.neurall.io/…/video.mp4" },
    "expiresAt": "2026-08-04T09:12:44.000Z"
  }
}

output.video is a signed CDN URL. Download it: anything the API makes is reaped 24 hours after the request.

Every status, the webhook signature, what a failure refunds and how long an output lives: How a generation works.

Prices you can fetch before you spend

Some features cost a fixed amount per call, others bill per second of output, and options like resolution move the rate. GET /v3/pricing returns your exact numbers, variants included, so a call can be costed before it is made. One prepaid balance for the whole organization, optional auto-recharge, automatic refunds when a generation fails.

Async by design

Creative renders take time, so every call is an async generation: submit, then poll or take a signed webhook, then download.

Recipes, not models

You call a feature, not a model name. Each feature runs a fixed, high-quality pipeline underneath, tuned for results you can rely on.

Built for agents

Every reference page is also served as markdown, the whole surface as llms-full.txt, and the contract as openapi.json.

Browse the docs

Start

Features

Images

Video

Audio

The API

Platform

Agent-readable docs live at neurall.io/llms-full.txt, and the app’s own reference at app.neurall.io/docs.