Updates·October 19, 2026, 04:51
Adobe lets Firefly users choose from more than 10 AI models
AI-generated from the sources listed, with automated source checks.
Adobe says its Firefly image app can now use more than 10 AI models from other companies, including Google, OpenAI and Runway. Adobe also promises that users' material will not be used to train the models.

Adobe says on its India product page that the AI tool Firefly supports more than 10 so-called partner models. These are AI models built by other companies that can be used directly inside Adobe's app.
The idea is that users should not have to switch between multiple apps. According to Adobe, Firefly is one place where you can pick the model that suits each step of a creative process, such as images, video or audio.
Which models are included?
Adobe names Google Gemini (Nano Banana), Google Veo, Google Imagen, OpenAI GPT Image 2, Runway Gen-4, Black Forest Labs Flux 1.1, Topaz Bloom, ElevenLabs and Kling AI, among others.
The Indian newspaper The Tribune also mentions Gemini 2.5 Flash Image, Flux.1 Kontext, Topaz Astra and Ideogram. The lists are not identical, and it is unclear exactly which versions are included. Adobe writes GPT Image 2, for example, while The Tribune simply writes GPT Image.
According to The Tribune, GPT Image is used for photorealism, Ideogram for text and typography, and ElevenLabs for speech. That is Adobe's angle and not an independent test. The newspaper's article appears under the heading "partner-exclusives," so it is probably sponsored or produced together with Adobe.
Promises on data and commercial use
Adobe says that neither Adobe nor its partners train models on users' personal, uploaded or generated material. Content also gets standardized information, called Content Credentials, showing who made it and how it was made.
Adobe claims that material created in Firefly and through the partner integrations can be safely used for commercial and marketing purposes. That is Adobe's own assessment.
The Tribune also writes that Firefly's own model, Firefly Image Model 5, is trained on material without copyright (public domain) and controlled datasets.
Not a brand-new idea
Adobe added the first third-party models to Firefly in the spring of 2025. These were OpenAI GPT Image, Google Imagen 3 and Veo 2, and Flux 1.1 Pro. Pika, Ideogram, fal.ai, Luma and Runway were to follow, according to the design outlet We and the Color.
The figure "more than 10" is Adobe's own count. Other sources from 2025-2026 talk about up to 30 models in Firefly, so the number depends on what is counted. Adobe's product page is localized for India and has no clear date.
Source
Get the week's AI news in your inbox
Choose your level, topics and length. One email a week, unsubscribe at any time.
Subscribe to Promptly Newsletter



