Adobe MAX 2026: Project Link Up turns dragging into context-aware vector fusion

Adobe used its MAX 2026 Sneaks session to demonstrate Project Link Up, a prototype that converts the ordinary drag gesture in vector design into context-aware fusion: when a designer places an object into a scene, the system adapts its perspective, scale and style to match the surroundings automatically. The chore it targets is concrete — manually redrawing an asset from a different viewing angle each time it enters a new composition.

The project comes from a team led by Mohit under Hailin Jin, senior principal scientist and research lead for imaging and design at Adobe. Jin frames the starting point as an observation about an everyday interaction.

"Design tools usually interpret that as an instruction about placement, but there's much more intent in what a designer moves, where they move it and how they move it," Jin said in an interview at MAX 2026.

Placing an illustrated car into a parking space, for example, normally requires separately adjusting its perspective, scale and style. The team asked whether the gesture itself could communicate more of that creative direction while keeping the result editable.

What control does the designer keep?

Adaptation is not an all-or-nothing transformation. The current prototype includes a creativity slider that guides how subtle or expressive the adaptation should be, ranging from preserving more of the original form to exploring a more dramatic transformation.

"The goal is to combine the ease of a gesture with explicit creative control," Jin explained. "The resulting vector artwork remains editable, so designers can continue refining it."

Editability is a structural constraint of the project, not an afterthought. Every output remains vector artwork that designers can continue to shape after the system has performed its adaptation.

How does it handle clashing styles?

Two fundamentally opposed vector elements do not have to merge into one. Jin describes the key as understanding the creator's intent: should the object match the scene, retain its own personality, or create a deliberate contrast?

Project Link Up combines three inputs to resolve that question:

• the drag gesture and its scene context
• creative controls such as the slider
• prompt guidance, for example "match the scene's illustration style" or "keep this character's style, but adapt its perspective and lighting"

The aim is to let designers steer the relationship between elements rather than accept a single automated result.

What is the philosophy behind it?

Jin's argument concerns the cost of exploration. When combining two elements requires substantial adjustment before a designer can judge whether an idea works, that effort interrupts discovery. Link Up lets designers try different relationships between objects with a familiar gesture, see the result immediately, and develop the ideas that hold up — in the way a quick physical sketch outpaces a week of careful drafting.

For daily practice, Jin identifies the single most time-consuming chore the tool could remove: manually reworking an asset to fit a new scene, particularly redrawing it from a different perspective. If adaptation happens as the designer places the object, less preparation is needed to evaluate a composition, leaving more time for the creative decisions that make it work.

Where does Adobe see this in five years?

A major direction in Adobe's research is helping tools understand creative content and the relationships within it. On that path, moving an object into a scene would communicate an intention about perspective or composition rather than simply a new position.

"For experienced designers, the opportunity is to express their judgment more directly and retain precise control over the result," Jin said. "We want to learn what creators expect a gesture to mean and how tools can respond in ways that feel useful and predictable."

Link Up is one part of that future. If the prototype matures into shipping software, the constant redrawing of objects at new viewing angles — a fixed cost of vector composition for decades — becomes a step the interface performs on the designer's behalf.