Learn what generative AI creates, when to choose raster or vector output, and how to turn a prompt into finished design work in Linearity.
Linearity Curve generates raster images and real vector artwork from a prompt. Vector results arrive as editable shapes, paths, and layers; raster results can be reprompted, resized, cropped, masked, and combined with the rest of your design.
Generation is available in Linearity on Mac, iPad, and the web, so you can create and continue refining your work across supported devices.
The foundation
AI-assisted design uses generative models to create a visual starting point from instructions, references, and constraints. The generated result is not the finished design: designers still direct the concept, correct the composition, apply brand decisions, and prepare each final format.
In Linearity Curve, the same workflow can begin with either a raster image or editable vector artwork. That distinction determines how the result behaves when you recolor it, reshape it, scale it, or reuse it across a campaign.
Choose an output
Choose based on what must remain editable, how the artwork will scale, and where the finished asset will be used.
Raster generation is suited to photographic scenes, textures, atmospheric artwork, and detailed visual exploration.
Vector generation is suited to icons, illustrations, graphic assets, and artwork that must remain scalable and editable.
From prompt to production
Reliable results come from a sequence of creative decisions, not from finding one perfect prompt.
Start with the audience, message, format, and intended action. A clear design brief gives generation a useful direction.
Explore multiple compositions instead of polishing the first result. Compare hierarchy, silhouette, mood, and suitability for the final format.
Reprompt raster imagery or refine vector shapes and layers. Correct structure first, then typography, colors, details, and brand decisions.
Test the design at every required size, verify readability and consistency, then export the correct files for web, social media, or print.
Product-specific guidance
Learn the workflow
Practical explanations for choosing image or vector output and turning generated material into finished design work.
Compare raster and vector output, understand the tradeoffs, and choose the right starting point for your project.
Learn how to prompt, reprompt, crop, mask, resize, and combine generated raster images inside a broader design.
Generate paths, shapes, and layers that stay editable, then refine their structure with familiar vector tools.
Go deeper
Connect AI generation with Linearity's existing vector tools, design principles, and production workflows.
See vector and image generation in the product, including prompts, generation modes, ratios, and editing workflows.
Understand the technology behind editable vector generation and how it fits into a practical design workflow.
Use Auto Trace when you already have a raster reference and need paths and shapes you can continue editing.
FAQ
An AI image is a grid of pixels, while an AI vector is constructed from mathematical paths and shapes. Images support photographic detail; vectors provide direct structural editing and resolution-independent scaling.
Yes. Vector results generated in Linearity Curve remain editable as paths, shapes, colors, and layers, so designers can restructure the artwork instead of accepting a flattened result.
No. Generation accelerates exploration, but designers still define the goal, choose the strongest direction, correct the composition, apply brand standards, and prepare usable final assets.
Curve supports production formats including SVG, PNG, JPEG, TIFF, PDF, and AI. Print work can also be prepared as a PDF with bleed where required.