Luma AI

Luma AI is a 3D Capture company founded in 2021 and based in Palo Alto, United States. It has raised $1.1B in total funding, most recently a Series C in 2025 at a $4B valuation.

Gaussian Splatting consumer leader — generates photorealistic 3D scenes from smartphone photos, plus the Dream Machine generative video model.

Sets the consumer quality bar for Gaussian Splatting; bridges AI-generated content into spatial computing pipelines.

lumalabs.ai 3D Capture

United Statesprivate

At a glance

Total raised

$1.1B

last round series c

Headcount

201-500

approximate · public range

Headquarters

Palo Alto

United States

Founded

2021

5 years active

Status

Private

3D Capture

Latest valuation

Funding history

Total raised

$1.1B

Rounds tracked

Latest stage

Series C

Latest valuation

$4B

Date	Stage	Amount	Valuation	Lead investors
Nov 19, 2025	Series C	$900M	$4B	HUMAIN
Dec 6, 2024	Series C	$90M	—

Founder backgrounds

Alberto Taiuti

Co-Founder

Attended Abertay University in Scotland, then worked at Codeplay Software before joining Apple as an AR/VR Engineer. Previously served as a Senior Autonomy Software Engineer at Skydio, working on drone vision systems. Co-founded Luma AI in August 2021 with Amit Jain and Alex Yu, where he served as CTO and built the infrastructure behind one of the world's most widely used 3D and video generation platforms. Published Slide AR and consulted on AR iOS projects. His technical background spans AR/VR engineering, computer vision for autonomous systems, and 3D graphics—expertise directly channeled into Luma's progression from NeRF-based 3D capture to generative video.

Alex Yu

Co-Founder & CTO

Graduated from UC Berkeley, where he conducted early research on NeRF neural rendering under Professor Angjoo Kanazawa, authoring pioneering papers on real-time neural rendering of 3D scenes and generation of 3D from a single image—including pixelNeRF and PlenOctrees, which rendered at more than 150 fps at 800x800px resolution, over 3000x faster than conventional NeRFs. Previously worked as a Research Intern at Adobe and also gained experience as a SWE Intern at Google. After graduating in 2021, turned down PhD offers from Stanford and MIT to start Luma with co-founder Amit Jain, applying his neural radiance field expertise to build generative models, in particular ray (aka Dream Machine) and genie, a text-to-3D model.

Amit Jain

Co-Founder & CEO

Earned a B.S. in Mathematics and Computer Science from Missouri Valley College, then spent four years at Apple as a Computer Vision, Cameras, and Systems Engineer, where he led development of the Passthrough feature for Apple Vision Pro and was instrumental in integrating the first LiDAR sensors into the iPhone. His background spans early work in physics and mathematical simulation through prior engineering roles at Circle Medical, Product Hunt, and founding Cultured Pixel. Co-founded Luma AI in 2021 to build what he describes as a "universal imagination engine". His public posture centers on creative empowerment: he frames Luma's AI as giving "humans their curiosity back" and offering artists "liberty to explore unconstrained" rather than optimizing industrial workflows.

AI bull / bear

Bull case

Luma AI has transformed from a NeRF and Gaussian-splat 3D-capture startup into one of the best-funded generative-media labs in the world. A $900 million Series C in November 2025, led by Saudi Arabia's HUMAIN, valued it north of $4 billion on roughly $1.1 billion raised, with a16z, Amazon, AMD, and NVIDIA on the cap table. Its Dream Machine platform claims more than 30 million users, and Ray3, launched in September 2025, arrived as what Luma calls the first reasoning video model, with native HDR output positioned against OpenAI's Sora 2 and Google's Veo 3\. Crucially for spatial computing, Luma's stated ambition is world models, AI systems that understand and simulate 3D space and physics, the same substrate embodied AI and immersive environments will demand. Backed by Project Halo, a planned two-gigawatt Saudi supercluster, plus Adobe and AWS partnerships, Luma is making one of the most aggressive compute-and-capital bets that generative 3D and video become foundational infrastructure.

Bear case

At more than $4 billion, Luma is priced for dominance in the single most brutal arena in AI. Generative video is a capital incinerator, and Luma competes head-on with OpenAI, Google, Runway, and Kuaishou's Kling, rivals with deeper compute, broader distribution, and in some cases stronger models. Its Dream Machine already faces copyright scrutiny over training data, a live legal and reputational exposure rather than a hypothetical one. The funding carries its own caution: heavy reliance on Saudi capital through HUMAIN concentrates strategic dependence on a single state backer, and world models remain scientifically unproven and ruinously expensive to train. The original 3D-capture business that earned Luma its spatial-computing relevance is now a secondary concern. If the video-model race compresses margins toward zero, the likely outcome when four giants are sprinting at once, a $4 billion valuation may prove a high-water mark rather than a floor.

Generated by editorial-sourced · May 29, 2026Refreshes quarterly · see disclosures

Products

Dream Machine
Dream Machine is Luma AI's flagship generative AI video platform that creates photorealistic videos from text, image, or video prompts. The platform serves over 30 million users globally and enables creative professionals to generate professional-grade video and images with natural, coherent motion. Users can input text descriptions, still images, or existing video clips to produce high-fidelity visual content for storytelling, advertising, animation, and gaming applications. The system learns from multimodal data including video, audio, and language to build world models capable of reasoning about motion, context, and physics.
Ray3
Ray3 is Luma AI's advanced reasoning video model launched in September 2025, representing the world's first reasoning video model capable of creating physically accurate videos in studio-grade HDR format. Ray3 features native HDR support with 10-bit, 12-bit, and 16-bit color depth, enabling production-ready fidelity, complex motion, character consistency, and built-in visual reasoning. The model can interpret visual annotations, allowing creatives to draw or scribble on images to direct motion and composition. Ray3 offers Draft Mode for rapid iteration and a Hi-Fi diffusion pass that delivers high-fidelity 4K HDR footage 5x faster and at lower cost.
Genie
Genie is Luma AI's text-to-3D generative model that enables users to create interactive 3D objects and assets from text prompts. The model generates complete 3D scenes with materials, quad mesh retopology, variable polycount, and standard export formats in under 10 seconds. Genie is designed for 3D game developers, product designers, and creative professionals who need to rapidly prototype and iterate on 3D concepts. The model supports photorealistic 3D capture using Gaussian Splatting technology, allowing users to transform smartphone photos into interactive 3D scenes.
Luma Agents
Luma Agents is Luma AI's agentic creative platform powered by its Unified Intelligence models, designed to coordinate multiple AI systems and generate end-to-end creative work across text, images, video, and audio. The agents maintain persistent context across assets, collaborators, and creative iterations, enabling users to steer creative direction through conversation rather than constant re-prompting. Luma Agents are used by global advertising agencies, brands, and studios to automate complex creative workflows, reducing production timelines and costs while maintaining quality standards and brand consistency.
Interactive Scenes
Interactive Scenes is Luma AI's photorealistic 3D capture and rendering platform utilizing Gaussian Splatting technology. It enables users to capture 3D content from smartphone photos or video and create browser and phone-friendly volumetric renders that stream efficiently (8-20MB files) while maintaining ultra-high quality. The platform supports real-time rendering on low-end devices and enables embedding of interactive 3D scenes anywhere on the web. Users can create lifelike 3D captures with natural camera moves and export renders in standard formats for commercial use across cinema, gaming, AR experiences, and web applications.

Luma AI

At a glance

Funding history

How they make money

Investability Index

Founder backgrounds

Alberto Taiuti

Alex Yu

Amit Jain

AI bull / bear

In the news

Leadership

Products

Dream Machine

Ray3

Genie

Luma Agents

Interactive Scenes

Open roles

Compare Luma AI with