AI & Machine Learning

Luma AI

4.38

builds 3D and video generation models, including Dream Machine for creating videos from text prompts.

Visit Website

Luma AI was founded in 2021 by Amit Jain and Jiaming Song in San Francisco. The company started with photorealistic 3D capture technology using neural radiance fields (NeRFs) and expanded into generative AI for video and 3D content.

Luma raised over $70 million, including a $43 million Series B in 2024. Investors include Andreessen Horowitz, GV, and Matrix Partners. The funding helped accelerate development of their multimodal generation capabilities.

The company’s Dream Machine, launched in mid-2024, generates short video clips from text or image prompts. It quickly gained attention for producing physically plausible motion and realistic camera movements, competing directly with Runway and Pika in the AI video generation space.

Luma’s earlier product, the Genie 3D model, generates textured 3D meshes from text descriptions. The company also offers a popular iOS app for 3D scanning that uses iPhone’s LiDAR sensor to create detailed 3D models of real objects and environments.

Luma AI targets creative professionals, game developers, and e-commerce teams who need to produce visual content efficiently. Their API allows developers to integrate 3D and video generation into their own applications. The company has around 50 employees and maintains a strong research presence, publishing work on neural rendering and generative models.