Luma

What is Luma Dream Machine?

Luma Dream Machine is the former public name for Luma’s generative media workspace. Although the Dream Machine name remains in legacy URLs and documentation, the current product is branded Luma. Its main purpose is to help users develop visual and audio content from prompts and reference files, then revise, organize, and share the resulting assets in projects or boards.

The product is most relevant to filmmakers, advertisers, marketers, designers, social-media creators, agencies, game developers, and businesses that need to explore concepts or produce short-form media without building every asset manually. It is closer to a multimodal creative production environment than to a general-purpose conversational assistant.

What Luma can create

Images and visual concepts

Luma can generate images from text prompts and reference images. This supports concept development, storyboards, campaign visuals, product ideas, and alternate creative directions. Users can iterate on an image rather than treating the first generation as a finished result.

Short videos from text or images

The platform supports text-to-video and image-to-video generation. A user can begin with a written description or a still image, select from available models and settings, and generate a short clip. Keyframes, continuation tools, camera controls, and related direction features help guide motion and transitions.

Video modification and reformatting

Luma is also designed for modifying existing or generated footage. Users can transform video, extend sequences, and reframe media for different aspect ratios or delivery formats. These functions make it useful for adapting an idea across social, advertising, and previsualization workflows, although output quality and consistency can vary by model, prompt, source material, and settings.

Audio for visual projects

Supported workflows include music, sound effects, and text-to-speech audio. Audio generation is an important extension of the visual workflow, but Luma should not be evaluated as a dedicated speech transcription or voice-cloning platform: the supplied product information does not identify speech-to-text or voice cloning as core capabilities.

How people use it in practice

A typical workflow starts in the web or iOS app. The user signs in, creates a board or project, enters a prompt, and optionally uploads an image or video reference. Luma generates an asset that can be revised, extended, reframed, downloaded, organized, or shared. This makes the product useful for moving from rough visual idea to a set of alternatives rather than simply producing one final file.

• Previsualization: Develop film, advertising, game, or product concepts before committing to a full production.
• Campaign development: Produce alternate images, short promotional clips, and format variations for different placements.
• Social content: Animate still images, create short videos, and adapt media to different aspect ratios.
• Creative iteration: Compare styles, camera directions, compositions, and visual treatments inside a project.
• Production workflows: Combine generated visuals with music, sound effects, or voiceovers and use the API or partner integrations where a more repeatable workflow is needed.

Models, platforms, and integrations

Luma’s current workspace can expose both Luma models and models from other providers. The supplied research lists Ray3.2, Ray3.14, UNI-1, Seedance 2.0, Kling 3.0, Kling Omni, Veo 3.1, Nano Banana, GPT Image 2, Seedream, ElevenLabs v3, ElevenLabs SFX v2, and ElevenLabs Music v1. Availability depends on the product surface and plan, so the model list should not be treated as a fixed promise for every account.

The main consumer surfaces are the browser-based Luma App and an iOS app. Luma also provides a developer API, and its broader ecosystem documents integrations or partnerships involving services such as Envato, Comfy, Runware, Flora, Krea, Magnific, Fal, and LovArt. These options matter for production teams that want to connect generation to an existing creative workflow, while ordinary users can work primarily within projects and shared boards.

For comparison, dedicated tools such as Runway, Midjourney, Leonardo.Ai, and Adobe Firefly occupy overlapping parts of the generative image and video market. Luma’s distinguishing emphasis is the combination of image, video, audio, reference-guided generation, media modification, and project-based iteration in one workspace. This does not mean it replaces a full traditional editor or every specialized production tool.

Pricing and access

The current public pricing structure is credit-based. Individual plans start with Plus at $30 per month, with Pro at $90 per month and Ultra at $300 per month. Annual billing is also offered at different effective prices, including $300 per year for Plus, $900 per year for Pro, and $3,000 per year for Ultra. Team and Enterprise access is available through separate offerings, with Enterprise pricing handled through a sales process.

Luma documents free trial credits, but the supplied current pricing information does not verify a continuing free subscription tier or a universal trial duration. Generation consumes credits according to factors such as the selected model, resolution, duration, audio options, and editing operation. Monthly credits expire at the end of the billing cycle and do not roll over. Users can enable additional usage under an optional spending limit, so high-volume generation can cost more than the base subscription.

Important limitations

• Credits can be consumed quickly: Longer, higher-resolution, or more complex generations may use substantially more credits, and unused monthly credits do not roll over.
• Results are variable: Video continuity, motion, subject consistency, and adherence to detailed instructions depend on the model and source material.
• Product branding has changed: Dream Machine is now a legacy name, and users may encounter older documentation, Ray terminology, and account information alongside the current Luma App.
• It is not a full conventional editor: Luma provides generation and modification tools, but users seeking comprehensive timeline editing or other specialized post-production controls may need additional software.
• Commercial terms vary: Rights and training-related treatment depend on the applicable plan and current terms. Users should verify licensing before using generated or uploaded material commercially.
• Platform coverage is limited or uncertain: Browser and iOS access are documented, while Android availability was not verified in the supplied current information.

Privacy and data considerations

Luma processes user inputs and outputs to provide the service. Enterprise customers can use contractual data-processing terms and organizational controls such as SSO, team management, shared credits, and usage analytics. The supplied information does not establish a universal retention period for active consumer content, although account-management documentation states that deleting an account permanently removes its videos, prompts, credits, and plan details.

Users should also pay attention to sharing settings. Shared boards or assets may be accessible to anyone with the link until they are unpublished. Historical Dream Machine licensing information indicates that content treatment and AI-training rights differ between lower and higher plan categories, and current plan names have changed. For confidential or commercially sensitive work, review the terms for the specific current plan rather than relying on older Dream Machine documentation.

Who should use Luma?

Luma is a good fit for creators and teams that need to explore visual ideas quickly, generate short image-based or text-based videos, revise footage, produce multiple aspect ratios, and keep related assets in an organized creative workspace. It is particularly useful when ideation and iteration matter as much as final rendering.

It is less suitable for someone seeking a free unlimited generator, predictable low-cost output at high volume, a dedicated research assistant, or a conventional video editor with complete professional post-production controls. Teams should also budget for credit usage and confirm commercial, privacy, and training terms before putting sensitive client material into the service.