Make your first API call.

1. Prepare your API key

Create a key in the customer console. Keep the secret in a server-side environment variable named SCENEPOND_API_KEY. The website login session is not an API key.

export SCENEPOND_API_KEY="YOUR_SCENEPOND_API_KEY"

2. Set your base URL

Send Authorization: Bearer with your API key. Use HTTPS and keep keys out of browser code and URLs.

3. Send a request

1curl "https://scenepond-api-production.up.railway.app/v1/chat/completions" \
2  -H "Authorization: Bearer $SCENEPOND_API_KEY" \
3  -H "Content-Type: application/json" \
4  --data-binary @- <<'JSON'
5{
6  "model": "deepseek-flash",
7  "messages": [
8    {
9      "role": "user",
10      "content": "Hello!"
11    }
12  ],
13  "stream": false
14}
15JSON

This example uses the chat completions protocol. Image, video, and task APIs can use different request schemas; check each model’s endpoint and configured provider before integrating.

Asynchronous tasks

Image and video task APIs return an ID before generation finishes. Submit once, preserve the ID, poll status, and retrieve media after successful completion.

Use the model page for exact required fields, optional generation settings, and Video or Responses compatibility. Video Responses require background:true; this does not apply to text-model Responses.

Use the same account API key to retrieve task results. Wait for SUCCESS or completed before downloading; on FAILURE or failed, inspect the returned error. Do not blindly repeat a paid POST after a timeout.

4. Monitor balance and usage

Requests are settled by the existing backend. Compare the configured model price before calling and review actual charges in your usage records.

Handle API errors

Model API documentation

Choose a model to view its API protocol and request examples.