What Is Adobe Firefly? Features, Free Access, and Commercial Use

Adobe Firefly is a generative AI service for creating and editing images, video, audio, and other content from text descriptions or reference images. You can use it to visualize ideas in a browser or work with generative features in Adobe apps such as Photoshop and Illustrator. Beyond generating a single image, Firefly helps you compare concepts and refine individual parts of a creative project.

What is Adobe Firefly?

Adobe Firefly is the name of Adobe's family of creative generative AI models and the web and mobile service that provides access to these tools. Generative AI creates new content based on descriptions or reference material supplied by a user. In Firefly, you can describe a scene, provide a reference image, and adjust settings such as composition and style.

The description or instruction you give an AI model is called a prompt. For example, “a small potted plant by a window in morning sunlight, watercolor illustration” describes both a subject and a mood. An image model uses that description to generate a result. The same prompt can produce different images, so comparing variations and refining them is part of the process.

The Firefly web app also offers partner models from companies such as Google and OpenAI. Using the Firefly service does not necessarily mean every result is generated by an Adobe model. Available models and controls depend on the feature and your account. See the official Adobe Firefly overview for the current range of tools.

What can you do with Firefly?

Start by deciding what you want to make. For a blog illustration, begin with image generation. For a short moving scene, explore video generation.

Generate images and edit selected areas

Image generation creates a new illustration or photographic image from scratch. Describe the subject, background, lighting, and style, then use the aspect ratio and reference controls supported by your chosen model. Aspect ratio is the relationship between an image's width and height. Choose a shape that suits its destination, such as a wide banner or a square post.

Generative editing changes an existing image. Generative Fill can add or replace content in a selected area, while Generative Expand can extend the scene beyond the original image boundaries. Distinguishing between creating a whole image and editing a specific area helps you choose the right tool. Adobe's guide to image generation and generative editing in Photoshop explains these workflows.

Reference images serve different purposes. A composition reference guides the arrangement and outlines of subjects, while a style reference guides the visual treatment, including color and texture. Think of references as a way to communicate a direction rather than a guarantee of an identical copy.

Generate video and audio

Video generation creates moving clips from a description of a scene and its motion, or from an image. A prompt might describe “mist drifting over a calm lake as the camera slowly moves forward,” combining subject movement with camera movement.

Begin by generating individual shots and editing them together rather than expecting one generation to produce a complete long video. Clip length, resolution, and available controls depend on the selected model and settings. Refer to the Firefly Generate Video FAQ for supported capabilities.

Audio tools include sound effects, speech and narration, and music generation. They can help you explore sounds that suit a video's mood, but check which models and features are available before choosing a workflow. When combining video and audio, adjust clip duration, sound timing, and volume during editing.

Compare concepts in Firefly Boards

Firefly Boards brings reference material and generated results together on one canvas. You can upload images and sketches, generate new images, and place alternatives side by side to compare design directions.

For a café promotion, for instance, a board could bring together concepts with warm natural light, muted colors, and minimal compositions. This is useful for a mood board: a collection of visual references that communicates colors, textures, composition, and atmosphere. Adobe's Firefly Boards documentation explains sharing, collaboration, and workflows with other Adobe apps.

How does Firefly relate to Photoshop and Illustrator?

Firefly is both a service you can use directly and a family of models powering generative features in Adobe apps. Examples include Generative Fill and Generative Expand in Photoshop and Generative Recolor in Illustrator. Illustrator works with vector graphics, which describe shapes using points and paths rather than a grid of pixels.

You might generate a concept or background in Firefly, refine colors and details in Photoshop, or place the image in an Adobe Express design to add text and finish the layout. A Firefly subscription does not automatically include every Adobe app. Check whether your plan includes access to the desktop applications you need.

Choose tools according to the task. The Firefly web app is a practical starting point for exploring concepts; a dedicated editing app is useful for detailed compositing or working with existing design files. Adobe's generative credits and supported features overview also describes generative tools across its apps.

Getting started: create an image from text

Image generation is a straightforward way to learn Firefly's basic workflow. You need an Adobe account and an internet connection, and you can use the web app in a browser. See Adobe's instructions for accessing Firefly if you need help signing in.

1. Open Firefly and sign in. Visit the Firefly web app and sign in with your Adobe account.
2. Open image generation. From the homepage, choose Image → Generate image. Menu placement and labels can change with updates or interface language.
3. Choose a model. Use the model selector under General settings. Check whether you are selecting an Adobe or partner model, whether it is available to your account, and its generation cost.
4. Set the shape and visual treatment. Choose an aspect ratio suited to the output, such as 16:9 for a wide banner or 1:1 for a square post. Use content type and reference controls when the model supports them.
5. Enter your prompt and generate. Describe the subject, background, composition, and lighting, then select Generate. The number of results varies by model.
6. Review and save the result. Check the composition and details, edit areas that need work or generate another variation, and download the image you want to use.

Adobe's text-to-image instructions explain the controls and generation history. If an option is missing, check your selected model first: controls available for one model may not be available for another.

How to write a useful prompt

Describe visible characteristics rather than relying on vague praise. Instead of asking only for a “great image,” explain what the image should show, where the scene takes place, and how it should be composed and styled.

• Subject: the main object, person, or landscape
• Setting: location, time of day, and surroundings
• Composition: close-up or overhead view, subject placement, and empty space
• Style: photography, watercolor, paper art, or another visual treatment
• Lighting and color: a specific mood, such as soft daylight and muted colors

For a wide blog background, you could start with a description like this:

A small potted plant and wooden desk beside a sunlit window, soft watercolor illustration, muted green and beige colors, generous empty space on the right for a title

If the result needs work, narrow down the issue instead of rewriting the entire prompt each time. Adjust the composition when you need more empty space, or the color and lighting when the image feels too intense. You can add a title or precise wording later in an editor such as Adobe Express or Photoshop. Adobe's guide to effective text prompts recommends concrete descriptions of subjects, style, and lighting.

Is Firefly free? Understanding generative credits

You can start using Firefly for free, but free access does not mean unlimited use of every feature and model. Free users receive limited daily generations on selected models and features. The allowance and available models can change.

For paid use, check the plan's generative credits. These are units used to measure consumption of generative features. The cost of a generation depends on the feature, model, and settings; generating an image and generating a video do not necessarily consume the same number of credits.

Depending on the plan, standard features may be unlimited without deducting credits, while premium features such as video generation and partner models use a monthly credit allowance. Compare which features your plan includes and which models consume credits, as well as the credit total. See Adobe's generative credits FAQ for usage rules and the Firefly plan comparison for prices and included apps. Choose your region when checking local pricing.

If you already subscribe to Creative Cloud, check your account's plan and credit balance before buying another subscription. The profile menu at the upper right of your Adobe account page shows generative credit information. Similar plan names do not necessarily provide the same features or allowances.

Can you use Firefly content commercially?

Adobe states that outputs from Firefly features without a beta label can be used in commercial projects. Outputs from beta features can also be used commercially unless the product explicitly states otherwise. Check the feature-specific notice and the commercial use guidance in Adobe's Firefly FAQ.

Adobe describes its Firefly training data as including licensed content, such as Adobe Stock, and public domain material. This description concerns Adobe's Firefly models; it should not be assumed to describe every partner model available in the app. Identify the model you use and the terms that apply. Adobe's partner model guidance explains the distinction.

Permission to use an output commercially does not automatically resolve every rights issue. Make sure you have permission to use uploaded references, and review results for recognizable trademarks or elements resembling someone else's content. For print work, check the downloaded file's actual resolution and how it looks at the intended print size.

Firefly also supports Content Credentials, metadata describing provenance such as creation and editing tools and AI use. Adobe says it applies these automatically to assets whose pixels are entirely generated by Firefly, including text-to-image outputs. See the Content Credentials overview for details about this information.

Can you submit Firefly images to Adobe Stock?

Images made with a commercially released version of Firefly can be submitted to Adobe Stock if they meet its generative AI submission guidelines and contain no watermark. Adobe Stock licenses assets such as photos, illustrations, and video. Uploading an image does not immediately make it available for licensing: it must pass moderation first. Adobe confirms Firefly eligibility in its Firefly FAQ for Stock contributors.

Check these conditions before choosing images to submit:

• Licensing rights: the terms of your AI tool and model must permit stock submission and commercial licensing. You also need the necessary rights to reference images and other input material.
• AI labeling: select Created using generative AI tools when submitting a Firefly-generated image. When it contains fictional people or property, also select People and Property are fictional as applicable.
• People and intellectual property: prepare the necessary releases for identifiable real people or recognizable real property. Do not include artist names, real people's names, fictional characters, or references to third-party intellectual property in prompts, titles, or keywords.
• Quality and originality: inspect hands, faces, and objects for generation errors and check technical requirements for the asset type. Avoid submitting multiple near-identical images from the same prompt.

See the Adobe Stock generative AI content guidelines for detailed requirements. Choose a content type that matches the result, such as Photos for suitable photorealistic images or Illustrations for artistic work. Generated content must not be submitted as an editorial record of a real event.

Prepare a contributor account in the Adobe Stock Contributor Portal, then follow this sequence:

1. Choose and upload images that meet the rights and quality requirements.
2. Select a content type and category, then enter a title and keywords describing the image. Do not use the generation prompt or technical settings as the title.
3. Confirm the AI label and any applicable fictional people or property label, and attach required releases.
4. Submit for moderation and check the decision. If an asset is refused, review the quality, technical, or legal feedback.

Adobe's AI content submission instructions explain the interface and process. Approval does not guarantee sales or income.

When is Firefly useful?

Firefly is useful for visualizing early ideas and comparing creative directions. You might make an illustration for a blog or presentation, explore a design's colors and mood, or develop scenes and sounds for a short video.

Actual photographs, screenshots, or diagrams can be more suitable when you need to show a product's exact appearance, explain numerical information, or document a procedure. Firefly creates new visual expressions, so distinguish generated material from product photography or evidence of real events.

Start with one image you need and compare results while adjusting prompts and settings within the free allowance. Once you know whether you will regularly use image, video, or audio tools, choosing the models and plan that fit your work becomes easier.