Luma Uni-1

• 🆕 Luma's unified image model combining reasoning and generation.
• 🧠 Understands user intent before producing finished visuals.
• 🎯 Designed for precise, iterative creative workflows.
• 🔧 Guides creation rather than one-shot prompting.
• 👁️ Turns text ideas into finished images.
• 🏢 Built by Luma AI, released June 2026.
• 📚 Documentation available via Luma's learning hub.

Luma AI

Luma AI is a research company focused on generative visual intelligence, building models that produce high-quality imagery and creative media. Its work centers on pushing the frontier of how machines understand and synthesize visual content, serving creators,…

Luma Uni-1 is described by Luma AI as a unified image model that turns ideas into finished visuals by combining reasoning with generation. Rather than treating prompting as a single pass, the company frames Uni-1 around understanding a creator's intent and guiding the image-making process toward a polished result.

A central theme in Luma's own materials is the emphasis on precise, iterative workflows: the model is positioned to interpret what a user wants, then help shape the output through successive refinement rather than relying on a fixed one-shot prompt. This "reasoning plus generation" framing distinguishes Uni-1 within Luma's image tooling.

Beyond these official descriptions, detailed specifications such as architecture, parameter counts, and licensing have not been disclosed in the sources reviewed here, so they are omitted. The model was released on June 17, 2026, under the catalog's authoritative date field.

Because no same-family predecessor is listed in this catalog entry, direct generational comparisons cannot be made responsibly. Readers seeking benchmark figures, supported resolutions, or API integration details should consult Luma's official learning hub and product documentation, which host the start guide and current capability descriptions for Uni-1.