Luma Launches Creative AI Agents Built on Unified Intelligence Models
Luma has unveiled Luma Agents, a new platform powered by its Unified Intelligence family of AI models, designed to handle end-to-end creative work across text, image, video, and audio. The launch marks a significant departure from current AI workflows that rely on stitching together multiple disconnected tools.
The system is built on the Uni-1 model, the first in Luma's Unified Intelligence family, which consolidates audio, video, image, language, and spatial reasoning into a single multimodal architecture. According to Luma CEO Amit Jain, the system enables "intelligence in pixels"—allowing agents to think in language and render directly into visual formats within a unified architecture.
How Luma Agents Work
Unlike traditional AI assistants that require users to piece together outputs from multiple models, Luma Agents operate as coordinated creative collaborators. The system can plan and generate content across multiple formats while integrating with external models including Google's Veo 3, ByteDance's Seedream, ElevenLabs voice models, and others.
Key capabilities include persistent context maintenance across assets, collaborators, and creative iterations. The agents perform iterative self-critique to refine outputs automatically and generate large sets of variations from simple text briefs. Users can steer creative direction through conversation rather than repetitive technical prompting.
Real-World Impact
The practical impact on creative workflows is substantial. In demonstrations, Luma showed how agents could transform a 200-word brief and an image into various ad campaign ideas. The company claims the system can localize a $15 million ad campaign for different countries in just 40 hours—a process that traditionally takes weeks.
Serviceplan Group, a major advertising network operating across 20+ countries, has already integrated Luma into its operations. The platform is pitched toward ad agencies, marketing teams, design studios, and enterprises looking to scale creative production without proportional headcount growth.
What Sets This Apart
The fundamental difference from current workflows is the unified multimodal reasoning system rather than assembling fragmented AI tools that require extensive prompting back and forth. This streamlines the creative process by reducing the need to master prompting across dozens of disparate models.
Luma Agents launched on March 5, 2026. Audio and video output expansions are planned for future Unified Intelligence releases.