List Sume's video endpoints from the OpenAPI JSON in Python

Sume publishes its API as an OpenAPI 3 JSON file, and the video endpoints are in it. A short Python script lists every /v1/videos and /v1/video-router operation so you can generate a client or diff against last week. The docs snapshot is at https://docs.sume.com/api/openapi.json; the live schema on the API service is the source of truth.

Which copy of the schema should you use?

The API reference says the markdown tables can be older than the live document, and names two places to read the schema: the local docs snapshot at /api/openapi.json and the live schema at https://api.sume.com/reference/json. Use the snapshot for a quick look and the live schema to generate code you will ship.

What does the script print?

It opens the snapshot and prints each operation whose path starts with the video prefixes.

import json, urllib.request

spec = json.load(urllib.request.urlopen("https://docs.sume.com/api/openapi.json", timeout=30))
for path, ops in spec["paths"].items():
    if path.startswith(("/v1/videos", "/v1/video-router")):
        for method, op in ops.items():
            print(method.upper(), path, "-", op.get("summary"))

What operations come back?

The snapshot lists these video operations.

What do you do with the list?

Feed the live schema to your generator of choice, then still call GET /v1/videos/models at run time. The schema describes the request shape; the model list says which ids, durations and reference types are open today. The video generation docs recommend /v1/videos for new integrations.