uni-1
Less Artificial. More Intelligent. uni-1 is Luma's first unified understanding and generation model — a multimodal reasoning AI that thinks before it generates. Create stunning images with structured reasoning, reference-guided precision, and unmatched creative control.
✨AI Image Generator Pro✨
uni-1 Advanced Capabilities
uni-1 combines multimodal reasoning with unified intelligence — a single model that understands and generates with unprecedented creative control.
Structured Reasoning for Image Synthesis
uni-1 performs structured internal reasoning before and during generation. It decomposes complex prompts, resolves spatial and logical constraints, and outscores competitors on reasoning-based benchmarks like RISEBench.
Unmatched Directability & Control
Take full control of your creative output with uni-1. Adjust camera angles, lighting, color temperature, and contrast. Use multi-turn image editing to refine results iteratively until they match your exact vision.
Style Transfer & Novel View Synthesis
uni-1 excels at style transfer and novel view synthesis, letting you reimagine scenes from new perspectives or apply artistic styles with precision. Create cinematic images, manga, and webtoons with a single unified model.
Multimodal Understanding
As a decoder-only autoregressive transformer, uni-1 processes text and images in a single interleaved sequence. This unified architecture enables true multimodal understanding — nearly matching Gemini 3 Pro on object detection while generating stunning visuals.