Adobe MAX 2026: Firefly Builder, New AI Tools and MCP

Adobe adds more AI to Photoshop, Premiere, Lightroom and Firefly at MAX 2026 and launches an MCP for ChatGPT, Claude and Gemini. Here's what's new.

Adobe is adding more generative AI to Photoshop, Premiere, Illustrator and InDesign. It announced the updates at its MAX event in Miami, along with expanded Firefly capabilities and a new Model Context Protocol (MCP) server. The MCP lets chatbots such as ChatGPT, Claude and Gemini use Adobe features directly. Taken together, the announcements show Adobe trying to make AI a working layer across its creative software, with Firefly as the hub.

This guide covers what Adobe announced, how the new tools work, and what they could mean for designers, video editors and photographers.

What Adobe announced at MAX in Miami

The announcements fall into four groups:

Firefly: a prototyping tool called Builder, on-brand graph creation, a more precise image editor, and an AI video assembly tool.

Adobe's creative AI agent: the agent is now proactive and always on.

App updates: new features in Photoshop, Premiere and Lightroom.

Models and integrations: two image models from Topaz Labs, plus a local MCP that exposes Creative Cloud features to other AI tools.

Each group is covered below.

Firefly Builder: a prototyping tool aimed at Figma

Builder is the announcement most likely to draw attention from outside Adobe's usual audience. It is a new prototyping tool inside the Firefly app, currently in private beta. Users can create designs that show off an idea, which puts Firefly in territory long held by Figma.

Prototyping sits between rough sketching and finished production design. A prototype lets a team show how an app screen or a web page might look and behave before anyone writes code or commits to a final layout. Adding this to Firefly signals that Adobe wants the app to cover the early, exploratory stage of a project as well as the polishing stage.

Builder is in private beta, so many details are still unknown. Adobe has not said how widely it will roll out or how closely it will match the depth of dedicated design tools. The direction is clear, though. Adobe wants designers to start ideas in Firefly rather than leave for a competing platform.

On-brand graphs and image editing that leaves the rest alone

Firefly also gets a tool for creating on-brand graphs. Anyone who has built a deck or report knows the problem: charts exported from a spreadsheet rarely match a company's colors, fonts and visual rules. A graph tool that follows brand guidelines could save marketing and communications teams a lot of manual cleanup.

The new image editing tool lets users change one part of an image without changing the other objects in it. Earlier generative editing tools often altered more than the user wanted, so a request to swap one element could shift the lighting or warp a nearby object. Targeted editing gives creators more control, and control is what separates a toy from a professional tool.

AI video assembly in Firefly

Adobe is also releasing an AI-powered video tool in Firefly. It assembles a video from clips and other elements, including images and sound effects. A creator can bring together raw material and let the tool build a first cut, then refine it by hand.

This matters for people who produce a lot of short-form content, such as social clips, product videos and quick explainers. The slowest part of that work is often the first assembly pass, and a tool that handles it could free up time for creative decisions. Professional editors will likely still turn to Premiere for detailed work, which fits Adobe's apparent plan: Firefly for speed, the flagship apps for depth.

A proactive, always-on creative agent

Adobe is making its creative AI agent proactive and always on. The goal is for the agent to gather more context about a user's work and be present for any action, instead of waiting for the user to open a chat window and ask for help.

Adobe is also making quick actions easier to reach. Examples include removing a background and generating a thumbnail for a video. These are small, repetitive jobs that eat time across a working week, and one-click versions of them suit a persistent agent well.

An always-on agent raises fair questions about how much context it collects and how users can control it. Adobe has not answered those in the announcement, so they are worth watching as the feature rolls out.

Photoshop, Premiere and Lightroom updates

The app-level updates continue a trend. Adobe had already added AI features across Photoshop, Express, Premiere and Illustrator earlier in the year, and MAX brought another round:

Photoshop gets a new relighting feature and sliders for prompt-based generative styles. Relighting lets a user change how light falls on a subject after the photo is taken. The style sliders give finer control over generated looks than a text prompt alone.

Premiere gets a new color grading system for professionals and an object removal feature for video. Removing an object from moving footage has been a long and tedious task, so automating it would be a big time saver.

Lightroom gets a way to correct blinking eyes and a blemish removal tool. Both address everyday problems in portrait and event photography.

Notably, several of these features are aimed at routine pain points. A blink in an otherwise perfect group shot or a distracting object in a clip are the kinds of problems that cost editors hours.

Third-party models and Topaz Labs

Adobe already offers many third-party models inside its products, from companies including Google, Runway, ElevenLabs, Black Forest Labs, OpenAI, Kling and ByteDance. That gives users a choice of engines for images, video and audio without leaving Adobe's apps.

Adobe also acquired Topaz Labs, a maker of image and video enhancement tools, earlier this year. Topaz is now releasing two image models. Wonder 4 handles upscaling and denoising, and ImageHDR turns standard images into HDR. Both address quality, which complements the generative features: one helps clean up and enlarge existing images, and the other expands the tonal range of ordinary photos.

Adobe's MCP for Creative Cloud

The biggest strategic move is the new Model Context Protocol server. MCP is a standard that lets AI assistants connect to outside tools and data. Adobe already has ChatGPT and Claude integrations that give users access to some of its product functions. The new release is a local MCP that exposes more Creative Cloud features.

According to Adobe, the initial set of endpoints and capabilities covers Illustrator, Photoshop, After Effects and Premiere. A user working in a chatbot such as ChatGPT, Claude or Gemini can call on those apps without switching windows, which changes where a creative workflow can begin.

Adobe vice president Deepa Subramaniam told TechCrunch that the company is still deciding what to expose and will lean on user feedback. She said Adobe is still working out where, when and how to draw the line, and that it will watch how people use the MCP and keep adding to it. She added that Adobe has no need or desire to limit what the endpoint can expose, and that the aim is to support the creative workflow however a person chooses to work.

That is a notable stance. Many software companies restrict what outside AI tools can do with their products. Adobe is taking an open approach, at least for now, and betting that broad access will keep creators working inside its ecosystem even when they start from another AI assistant.

Figma's skepticism about MCP

Not everyone is convinced MCP is the right foundation. Last month, Figma co-founder Dylan Field said at an event in India that MCP is not a perfect standard and that it benefits frontier AI labs more than others. Field is the head of one of Adobe's main rivals, and his comments reflect an open debate in the industry over who gains most from the protocol.

The argument runs like this. If AI assistants become the main way people reach software, the companies that run those assistants gain a lot of influence, while the makers of the underlying tools risk becoming plug-ins. Adobe's decision to open its apps through MCP suggests it accepts that risk, or sees the upside as larger.

What this means for creators and teams

For working designers, editors and photographers, a few points stand out:

Faster routine work. Features like object removal, blemish removal, blink correction and quick actions target repetitive tasks.

More control. Selective image editing and prompt-based style sliders suggest Adobe is listening to complaints about unpredictable AI results.

New starting points. With the MCP, a project can begin in a chatbot and move into Photoshop, Illustrator, After Effects or Premiere.

A broader Firefly. Builder, graph creation and video assembly turn Firefly from an image generator into a wider creative workspace.

Teams should also test before relying on any of this. Builder is a private beta, the MCP is new, and Adobe says it is still deciding where its limits belong.

Frequently asked questions

What did Adobe announce at MAX 2026?

Adobe announced more AI features across Photoshop, Premiere, Illustrator, InDesign, Lightroom and Firefly. It also launched a local MCP so that chatbots can use Creative Cloud features.

What is Adobe Firefly Builder?

Builder is a prototyping tool inside the Firefly app. It is in private beta and lets users create designs that show off ideas, which puts it in competition with Figma.

What is an MCP, and why does Adobe's matter?

MCP stands for Model Context Protocol. It lets AI assistants connect to outside software. Adobe's MCP lets chatbots like ChatGPT, Claude and Gemini use features from Illustrator, Photoshop, After Effects and Premiere.

Which Adobe apps does the MCP cover?

The initial endpoints cover Illustrator, Photoshop, After Effects and Premiere. Adobe says it plans to keep adding more based on how people use it.

What new features are coming to Photoshop?

Photoshop gets a relighting feature and sliders for prompt-based generative styles.

What is new in Premiere?

Premiere gets a color grading system for professionals and an object removal feature for video.

What is new in Lightroom?

Lightroom gets a tool to correct blinking eyes and a blemish removal tool.

What does the Firefly video tool do?

It assembles a video from clips and other elements such as images and sound effects.

What does it mean that the Adobe agent is always on?

Adobe is making its creative AI agent proactive so it can gather more context about a user's work and be present for any action. It is also easing quick actions such as background removal and video thumbnail generation.

Which third-party AI models does Adobe offer?

Adobe offers models from Google, Runway, ElevenLabs, Black Forest Labs, OpenAI, Kling and ByteDance.

What are Wonder 4 and ImageHDR?

They are two image models from Topaz Labs, which Adobe acquired. Wonder 4 handles upscaling and denoising, and ImageHDR turns standard images into HDR.

Why is Figma's CEO skeptical of MCP?

Figma co-founder Dylan Field said MCP is not a perfect standard and benefits frontier labs more than others.

Reference

TechCrunch - https://techcrunch.com/2026/11/10/adobe-adds-more-ai-to-its-products-and-launches-an-mcp/

Adobe newsroom - https://news.adobe.com/news/2026/11/adobe-powers-new-era-of-creativity

Loaded All Posts Not found any posts VIEW ALL Readmore Reply Cancel reply Delete By Home PAGES POSTS View All RECOMMENDED FOR YOU LABEL ARCHIVE SEARCH ALL POSTS Not found any post match with your request Back Home Sunday Monday Tuesday Wednesday Thursday Friday Saturday Sun Mon Tue Wed Thu Fri Sat January February March April May June July August September October November December Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec just now 1 minute ago $$1$$ minutes ago 1 hour ago $$1$$ hours ago Yesterday $$1$$ days ago $$1$$ weeks ago more than 5 weeks ago Followers Follow THIS PREMIUM CONTENT IS LOCKED STEP 1: Share to a social network STEP 2: Click the link on your social network Copy All Code Select All Code All codes were copied to your clipboard Can not copy the codes / texts, please press [CTRL]+[C] (or CMD+C with Mac) to copy Table of Content