HappyHorse 1.0 API: what Runway lists and how to check Sume's models

Runway Dev lists HappyHorse 1.0 (happyhorse_1_0) for text or first-frame image to video, 3–15 seconds, at 720p and 1080p. Sume's video docs do not name a HappyHorse id, so do not assume one: read GET /v1/videos/models and use what it returns.

Sources: the Runway Dev changelog (May 29, 2026 entry) and Sume's video generation docs, read 2026-10-01.

What does Runway say about HappyHorse 1.0?

The entry says text-to-video supports 10 output dimensions across 720p and 1080p, and image-to-video preserves your input aspect ratio and accepts an optional motion prompt. It gives no price in that entry.

How do I check whether Sume has a model?

Call the model list and read each model's supported_resolutions, supported_aspect_ratios and supported_durations. The docs say limits are not uniform: seedance-2.5 accepts 4–30 seconds and wan-3.0 2–30, while models not named there top out at 15.

const res = await fetch("https://api.sume.com/v1/videos/models", {
  headers: { Authorization: "Bearer " + process.env.SUME_API_KEY },
});
console.log(await res.json());

How do Runway's and Sume's sizing fields differ?

Sume uses resolution plus aspect_ratio; every v1 model reports supported_sizes: null, so a pixel size is rejected.

What should I do next?

If you need a 3-second clip, check supported_durations first, because a model's minimum may be higher. Treat the list response as the source of truth, not a changelog from another vendor.