Luma Launches Uni-1.1 Image API Matching OpenAI, Google
Luma Labs now offers its top image generation model, Uni-1.1, through a developer API. This REST interface allows users to create and modify images using text prompts. On the Arena leaderboard, Uni-1.1 places 7th or 8th in both main categories. It trails just behind models from OpenAI, Google, and Grok Imagine.
Luma Labs, a company focused on generative AI tools, built its reputation with video generation products like Dream Machine. That tool, released in 2024, lets users produce short video clips from text or images. The firm, started by former Meta engineers in 2023, has quickly advanced in multimodal AI. Now, with Uni-1.1, Luma enters the competitive image generation space dominated by big players.
API Pricing and Tiers
The API comes in two versions. The basic "uni-1" tier costs $0.0404 per image generated. For better results, "uni-1-max" runs at $0.10 per image. Every output reaches 2,048-pixel resolution, suitable for high-detail work. Developers can add up to nine reference images per request. Each reference adds $0.003 to the cost. These rates put Luma on par with Google and OpenAI offerings. Full pricing and usage details appear in Luma's documentation.
This structure makes the API accessible for various projects, from quick prototypes to professional designs. Reference images help guide the model toward specific styles or compositions, a common need in creative applications.
Performance on Leaderboards
Arena serves as a key benchmark for AI image models. Users vote on outputs from different systems in blind tests. Uni-1.1's 7-8 ranking shows strong performance across text-to-image and editing tasks. It competes closely with top systems from OpenAI's DALL-E series, Google's Imagen, and xAI's Grok Imagine. Such placements indicate reliable quality for real-world use.
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
Luma's model handles complex prompts well, producing coherent and detailed results. The close competition highlights how quickly the field evolves, with new entrants challenging established names.
Features and Web Comparison
The API provides the same capabilities as Luma's web demo for Uni-1.1. That online version drew positive attention for its output quality. Both support deep text and image comprehension. They include reasoning steps to refine generations. The web tool uses an agent system for tasks like creating multiple image variants. For example, a prompt such as "make 20 variations of this image" works smoothly there.
Developers gain flexibility with the API. They can create similar iterative processes in their own applications. This suits custom workflows in apps or services. The agent approach in the web version handles back-and-forth refinements automatically.
Future Availability Plans
Luma plans to expand access beyond its direct API. Integration with cloud providers like AWS is in progress. No specific launch date has been announced. This move would simplify deployment for large-scale users. For now, developers access the service through Luma's platform.
The release positions Luma as a serious option in image AI. With matching prices and near-top quality, it appeals to builders seeking alternatives to the biggest providers.