Agent skill · Testing & QA

luma-ray-video

Direct and operate Luma Ray video generation through the current Luma Agents API and distinguish it from the consumer Luma App and legacy Dream Machine API. Use for Ray 3.2 text/image/multi-keyframe video, edits, extends, reframes, HDR/EXR, pricing and approvals, async delivery, migration, rights, privacy, and production QA.

Calesthio43,316★ · +2,384/wk · 2 repos on radarProfile →
claude-codecodexcopilotcursorMIT
Install
npx skills add calesthio/generative-media-skills --skill luma-ray-video --agent claude-code

Same command for any agent — swap --agent for codex, cursor, copilot.

Facts
Files in the skill folder: 2
SKILL.md size: 21 KB
Bundled scripts: none
Path: skills/providers/video-generation/luma-ray-video/SKILL.md
Open the folder on GitHub →
Where it comes from
Stars: 112 · +8 this week
Language: Python
Read our review of the source →

Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.

Review
written from the skill's own SKILL.md · Aug 5, 2026

What it does

Directs and operates Luma Ray video generation through the Luma Agents API (ray-3.2) for text/video, image-to-video, edits, reframes, HDR/EXR, pricing and approvals, async delivery, migration, rights, privacy, and production QA. It emphasizes using the Agents API surface for new code and treats the legacy Dream Machine API as a compatibility layer. It differentiates three surfaces (Agents API, Luma App, Dream Machine) and specifies prompts, output formats, and pricing rules.

How it works

  • Uses the current developer surface: Luma Agents API at base URL https://agents.lumalabs.ai/v1 with Bearer token LUMA_AGENTS_API_KEY and model ray-3.2.
  • Supports generation types: video, video_edit, video_reframe via POST /v1/generations; polls via GET /v1/generations/{id}.
  • Distinguishes from consumer app (Luma App) and legacy Dream Machine API; uses Luma Agents API for new code; treats Dream Machine as legacy compatibility surface.
  • Defines capabilities and constraints for Ray 3.2: text-to-video, image-to-video with anchors or multi-keyframes, extend, and reframe; specifies frame durations, resolutions, and aspect ratios (e.g., 9:16, 3:4, 1:1, 4:3, 16:9, 21:9) and output MP4s with selectable resolutions (360p–1080p).
  • For edits, requires exactly one source (generation_id, file_id, or URL/base64 with media_type) with per-channel duration limits; starts with video.edit.auto_controls: true and provides nine strength controls or per-signal controls (pose, depth, normals, trajectory, face), noting mutual exclusivity with auto_controls.
  • For reframing, requires a source and target aspect_ratio; SDR only; per-source limits and 200 MB per source for URL/data.
  • Includes pricing guidance and the requirement to create an approval record before paid POSTs, with explicit fields (canonical request, model/type, reference hashes, price, timestamp, max USD, rights, retention, approver, expiry), and failure modes on invalid/insufficient approvals.
  • Provides example request bodies demonstrating video, keyframes, and edits, including the use of video properties like duration, resolution, HDR, and end_frame/start_frame semantics.

When to use it

Use for Ray 3.2 video generation needs (text/video, image/video, edits, reframing) when working with the Luma Agents API (ray-3.2). Before any paid POST, ensure an approval record exists with the canonical request and price, and that the approval maximum is sufficient.

What it can touch

  • API: https://agents.lumalabs.ai/v1
  • Model: ray-3.2
  • Generation types: video, video_edit, video_reframe (via POST /v1/generations)
  • Polling: GET /v1/generations/{id}
  • Requires: LUMA_AGENTS_API_KEY sent as Bearer token; supports headers and request bodies shown in examples

Caveats

  • Volatile facts verified on 2026-07-10; recheck model page, schema, price, and limits before paid runs.
  • Legacy Dream Machine API remains documented; do not send legacy fields to the Agents API.
  • Pricing figures are from 2026-07-10 and may change; recheck before execution.
  • HDR/EXR and certain features have specific constraints (e.g., HDR requires 720p/1080p and 5 s for type video; EXR only when exr_export is true).
  • Approvals must be complete before paid POSTs; the dry-run example shows idempotency and tracing behavior but is not a guarantee of success.
From the SKILL.md

# Luma Ray video production Route the job before writing a request. Luma currently exposes similarly named products with incompatible models, schemas, credentials, prices, and data terms. For new API integrations, use the **Luma Agents API** and `ray-3.2`; treat the older Dream Machine API as a legacy compatibility surface. Volatile facts were verified from first-party Luma sources on **2026-07-10

More from generative-media-skills
All skills →
About this skill
What does the luma-ray-video skill do?

Direct and operate Luma Ray video generation through the current Luma Agents API and distinguish it from the consumer Luma App and legacy Dream Machine API. Use for Ray 3.2 text/image/multi-keyframe video, edits, extends, reframes, HDR/EXR, pricing and approvals, async delivery, migration, rights, privacy, and production QA.

How do I install it?

Run `npx skills add calesthio/generative-media-skills --skill luma-ray-video --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.

Where does this skill come from?

From calesthio/generative-media-skills, a repository with 112 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.

Is a popular skill a good skill?

Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.

Keep going