Adobe Bets Next Year on On-Device AI Partnerships

Adobe will spend "far more time" on on-device AI partnerships next year, according to Shilpa Anant, the company's vice president of global strategic partnerships. She made the comments on the 11th, during a media interview on day two of Adobe MAX 2026 in Miami.

On-device AI runs inside a computer's own CPU and GPU without touching a server. Because the two processors share memory, models execute locally — and skipping the server speeds up image and video work. That's the practical payoff for creatives.

What is Adobe's starting position?

Adobe hasn't been idle. The company has already embedded more than 30 external AI models into its services while linking flagship product functions to general-purpose AI chatbots. "We are doing a lot of good work in this area," Anant said.

The agentic AI features built so far are groundwork. On-device AI is where the next round of competition plays out, and Adobe wants a seat at that table.

Which chips matter here?

Two hardware threads run through Anant's argument:

• Nvidia's RTX Spark, a chip for Windows PCs unveiled this year. "You would have seen devices equipped with RTX Spark at the Nvidia booth set up at the MAX venue," Anant said.
• Apple's M-series, the in-house silicon where "Apple is also innovating to draw more GPU performance out of devices."

Adobe is already working with Nvidia to optimize its apps for laptops based on RTX Spark. At this year's MAX, it also announced an optimization effort with Microsoft on on-device models.

"The partnership with Microsoft is ultimately about delivering on-device AI," Anant said. Topaz Labs, which Adobe recently acquired, brings its own on-device models to the stack. Anant declined to name specific formal partnership plans beyond that.

She framed the goal simply: improve the creative experience wherever customers are, and let everyone create.

Why are partnerships moving so fast?

No single company can hold models, chips and services all at once. That constraint is forcing alliances across the AI industry to form faster than the usual corporate pace.

"In the past we would plan and prepare a partnership over several months, but now it comes together in weeks, sometimes days," Anant said. "In more than 20 years of doing this work, I have never seen this speed."

She added that top executives on the partner side now get involved down to the details — collaboration is happening very actively across the industry.

What else landed at MAX 2026?

Beyond Microsoft, Adobe announced expanded partnerships with Disney and Google. The Disney deal is the most concrete: the two companies agreed to develop AI models tailored to Disney's intellectual property using Firefly Foundry, Adobe's enterprise service.

Nvidia, for its part, sees the Adobe relationship as directly practical. An Nvidia official at a demo booth on the MAX venue said applying a toolkit such as RTX to software makes AI features run far more efficiently, which ultimately saves creators time.

"That is exactly where both companies find value in the partnership," the official said.

What changes for practitioners?

For creative professionals, the direction is clear. More Adobe tools tuned to local silicon — RTX Spark laptops, M-series Macs — means faster rendering and AI assistance that doesn't round-trip to a server. The missing piece is specifics: no formal on-device partnership roadmap was disclosed at MAX. Watch the Nvidia and Microsoft optimization work as the first concrete deliverables.