AI Usage at a Glance
Timeline
Jan 1, 2024
ModerationPractice documented: Luma AI uses a multi-layered moderation system that combines AI filters with human oversight to screen prompts and generated videos on Dream Machine and the Dream Machine API for policy violations.
Practice DocumentedView practice →Jan 1, 2024
Creative GenPractice documented: Luma AI offers a 3D capture mobile app that uses AI and Neural Radiance Field (NeRF) technology to turn smartphone video footage into interactive, photorealistic 3D scenes that can be shared on the web or exported to game engines.
Practice DocumentedView practice →Jun 1, 2024
Creative GenPractice documented: Luma AI offers Dream Machine, an AI-powered platform that generates short cinematic video clips from text prompts or uploaded still images. Users worldwide can access it via web, iOS app, or API under free and paid subscription tiers.
Practice DocumentedView practice →Nov 20, 2024
Creative GenPractice documented: Luma AI offers Modify Video (also called video-to-video), a feature within Dream Machine that lets users transform existing video clips by changing environments, visual styles, character appearances, or other elements using text instructions — without reshooting footage.
Practice DocumentedView practice →Dec 1, 2024
Creative GenPractice documented: Luma AI offers Photon and Photon Flash, a pair of AI image generation models that convert text prompts or reference images into high-resolution still images at up to 1080p. These models are available via the Dream Machine platform and the Luma API.
Practice DocumentedView practice →Jan 1, 2025
Creative GenNew evidence: Video to Video AI with Dream Machine for Pro-Level Edits | Luma
Evidence AddedView practice →Jan 1, 2025
ModerationNew evidence: Content Moderation - Dream Machine Guide
Evidence AddedView practice →Mar 1, 2025
Creative GenNew evidence: News & Updates | Latest Innovations & Releases | Luma AI
Evidence AddedView practice →Apr 24, 2025
OtherPractice documented: Luma AI integrated its Ray video generation models into Adobe Firefly, making them available to Adobe's global user base for text-to-video and video editing within Firefly's web app, mobile app, and Firefly Boards. Ray2 launched in Firefly in June 2025, followed by Ray3 in September 2025.
Practice DocumentedView practice →Sep 1, 2025
OtherNew evidence: Luma AI's Ray3 Now in Adobe Firefly | Adobe
Evidence AddedView practice →Sep 1, 2025
Creative GenNew evidence: AI Video Generation with Ray3 & Dream Machine | Luma
Evidence AddedView practice →Sep 23, 2025
OtherNew evidence: Adobe Adds Luma AI's Ray3 AI Video Generator to Firefly
Evidence AddedView practice →Jan 1, 2026
Creative GenNew evidence: Luma 3D Capture - App Store - Apple
Evidence AddedView practice →Feb 13, 2026
Creative GenNew evidence: What Is Luma Photon 1? Photorealistic AI Image Generation from Luma Labs
Evidence AddedView practice →Mar 1, 2026
Creative GenPractice documented: Luma AI deployed Uni-1, a unified AI model that combines visual reasoning and image generation in a single architecture, enabling users to generate new images or precisely edit existing ones using natural-language instructions. It became publicly available on March 23, 2026.
Practice DocumentedView practice →Mar 24, 2026
Creative GenNew evidence: Luma AI launches Uni-1, a model that outscores Google and OpenAI while costing up to 30 percent less
Evidence AddedView practice →May 1, 2026
Creative GenNew evidence: Introducing the Uni-1.1 API: Intelligence You Can Direct. Aesthetic You Can Ship.
Evidence AddedView practice →Recent activity
Luma AI offers Modify Video (also called video-to-video), a feature within Dream Machine that lets users transform existing video clips by changing environments, visual styles, character appearances, or other elements using text instructions — without reshooting footage.
Modify Video was introduced as an API capability in November 2024 and is available inside Dream Machine. It accepts an existing video clip plus natural-language instructions as input and outputs a transformed video that preserves the original motion while applying the requested changes. The Ray3 version (Ray3 Modify) adds precise keyframe and character reference controls, allowing changes such as relighting scenes, swapping environments, and altering costumes without green screens or post-production compositing.
Luma AI offers Dream Machine, an AI-powered platform that generates short cinematic video clips from text prompts or uploaded still images. Users worldwide can access it via web, iOS app, or API under free and paid subscription tiers.
Dream Machine was launched publicly on June 12, 2024. It accepts text prompts or still images as input and produces video clips with coherent motion and cinematic effects. The platform is powered by Luma's Ray model family (Ray2, then Ray3 as of September 2025, and Ray3.14 as a further update), which supports resolutions up to 1080p native, HDR output, and up to 10-second clips. A free tier offers 30 generations per month; paid plans (Lite, Standard, Pro, Premier) offer higher generation limits and commercial use rights.
Luma AI uses a multi-layered moderation system that combines AI filters with human oversight to screen prompts and generated videos on Dream Machine and the Dream Machine API for policy violations.
According to Luma's own API documentation, the moderation system applies to all video generations on the platform and via the API, covering categories such as harmful content, graphic violence, and sexually explicit material. API customers can customize the moderation settings to match their market and user base. Appeals for incorrectly flagged content can be submitted to Luma's moderation team via email. The system is described as combining automated AI filters with human review.
All practices by category
Luma AI offers Modify Video (also called video-to-video), a feature within Dream Machine that lets users transform existing video clips by changing environments, visual styles, character appearances, or other elements using text instructions — without reshooting footage.
Luma AI offers Dream Machine, an AI-powered platform that generates short cinematic video clips from text prompts or uploaded still images. Users worldwide can access it via web, iOS app, or API under free and paid subscription tiers.
Luma AI offers a 3D capture mobile app that uses AI and Neural Radiance Field (NeRF) technology to turn smartphone video footage into interactive, photorealistic 3D scenes that can be shared on the web or exported to game engines.
Luma AI deployed Uni-1, a unified AI model that combines visual reasoning and image generation in a single architecture, enabling users to generate new images or precisely edit existing ones using natural-language instructions. It became publicly available on March 23, 2026.
Luma AI offers Photon and Photon Flash, a pair of AI image generation models that convert text prompts or reference images into high-resolution still images at up to 1080p. These models are available via the Dream Machine platform and the Luma API.
Cite this record
Trace Foundation. (2026). Luma AI: documented AI practices (data as of 2026-05-11) [Data set record]. AI Trace. https://www.aitrace.org/r/company/f8893779-2bd9-4689-b231-ee85efcf9f2e. Accessed September 7, 2026.
- Stable link
- https://www.aitrace.org/r/company/f8893779-2bd9-4689-b231-ee85efcf9f2e
- Data as of
- May 11, 2026
- Last verified
- Not recorded
Related Companies
Trace Foundation
Activision Blizzard
Walmart
Stability AI
Topaz Labs
Have evidence about Luma AI's AI practices? Submit a report.
Submit a report →AI Trace is free and nonprofit. Support our work