Adobe has made a range of its creative and productivity tools available inside ChatGPT through a connector called Adobe for ChatGPT, allowing users to edit images, adapt videos and build documents by describing what they want in plain language.
The connector brings capabilities from Photoshop, Premiere, Firefly, Express, Acrobat and other Adobe applications into the chat window. Rather than switching between applications, users describe the task — retouching a batch of photographs, resizing a video, turning a spreadsheet into a set of printable badges — and the relevant Adobe tools carry it out on images, videos, documents and other files.
The first is photo editing at volume. Users can upload one or more images and describe the look they are after, with the connector adjusting colour, lighting and tone. It can also remove distractions, generate new elements, replace backgrounds, apply presets, blur, crop, resize and expand images. Adobe cites the example of uploading a set of team headshots and asking for the lighting to be balanced and the images cropped consistently for a company profile page. Users can also search their Creative Cloud folders for particular images — all beach photographs, say — and apply a consistent visual style across the batch.
The second is video. A horizontal video can be uploaded and adapted for different channels and formats, such as YouTube Shorts or Instagram Reels. The connector can also assemble a quick-cut video pulling out the strongest moments from a longer tutorial or walkthrough.
Third is template-based design and PDF creation. Users can begin with an Adobe Express template, change the text and colours, and then either animate the result for social media or convert it into a PDF. Adobe’s examples include an Instagram Story for a boutique sale and a flyer for a yoga studio that is customised and then converted into a print-ready PDF.
The fourth category covers data-driven documents. Users describe the output they want and supply the data, and the connector turns raw information into formatted, shareable PDFs — event badges generated from a list of attendee names, or cards, catalogues and structured documents built from a dataset.
Access to ChatGPT is required. An Adobe account is not needed to begin, though signing in or creating a free Adobe account unlocks additional skills, more generative AI tool calls, storage for assets and continuity across sessions.
To set it up, users open ChatGPT, select Plugins from the sidebar, search for Adobe and select Connect. A prompt then starts a workflow; those unsure where to begin can type “What tools are available in Adobe?” to see the options. Further prompts apply additional generative edits or adjustments.
Adobe has said the connector will continue to evolve as it expands the range of supported workflows and takes on user feedback.Adobe has also published the specifications it recommends for running the connector. On the browser side, macOS users are advised to be on Google Chrome version 143 or later, or Apple Safari version 26 or later, while Windows users should be on Chrome version 143 or later, or Microsoft Edge version 143 or later. The recommended operating systems are macOS 18.2 or later and Windows 10 or later. Adobe also recommends a 64-bit processor with four or more cores, and a stable internet connection is required.
For Chromebook users, the recommendation is Google Chrome version 143 or later, 8GB of RAM or more, and an Intel Core i3 processor or better, again with a stable internet connection.
