An on-brand AI image generator. Make the image you have in mind.

DFIRST Create image is an AI tool that turns a written description into a new picture. Describe the idea, add up to four reference photos of a product, a person or a look the picture should follow, choose 1:1, 16:9 or 9:16, and you get one image saved in your workspace.

Words in, one picture out
Up to four reference photos
Square, wide or vertical

Works with

Pictures with a point of view.

Made in DFIRST for the example brand ZESTA, each from a short description. With Brand DNA and References in the workspace, the colours, the products and the faces stay yours.

A woman laughing and holding a tangerine can toward the camera — “A woman laughing, holding a tangerine can toward the camera, shot from below, blue summer sky.”

Three ZESTA cans with citrus and sunglasses on orange — “Top-down flat lay on tangerine orange: three cans, citrus halves, white sunglasses.”

A skater drinking from a lime can at golden hour — “A skater on the lip of a bowl at golden hour, drinking from a lime can.”

Three basketball players lying on an orange court holding cans — “From directly above: three players lying on an orange court, cans raised.”

A grapefruit can in the sand in front of a beach volleyball game — “A grapefruit can in the sand, a beach volleyball spike behind it.”

A woman riding a bike with a can in the basket by the sea — “A woman on a city bike by the sea, a can and lemons in the basket.”

How to create an on-brand image with DFIRST.

Three steps. The first one starts on this page.

01
Describe the picture
Starts on this page
Write what is in the frame, where it is and how it should feel. Add reference photos if the picture should follow a product, a face or a style, and pick the format.
02
DFIRST draws it
DFIRST does the work
Your words and any reference photos go to the image model together. The references steer what the product looks like or how the light falls, and the words decide the rest.
03
Use it or keep going
Ready for you
Download the image or ask for a new take. On the board of a session you can switch the model, the format or the quality of the card and build the next picture from this one.

Made for your brand. Not a blank prompt box.

References, not guesswork

An AI image generator with reference images stays closer to what you mean. Show it your can, a model's face or a photo whose light you love, and the result follows it instead of a stock idea of your category.

Your saved cast, one click away

Pick pictures straight from References instead of uploading them again: the people, products and places your brand keeps coming back to.

Not tied to one engine

On the board of a session, the same image card can be redrawn with another model or a higher quality setting, so you can compare styles before you choose.

Who uses it.

Social media managers

Post images in the brand's look, made on the day they are needed.

Brand designers

Concept images and key visual ideas before anything is shot or illustrated.

Content marketers

Header images for articles, newsletters and landing pages in the right shape.

Agencies

Quick visuals for a pitch, built around the client's own product photos.

At a glance.

The facts about Create image, as it works in DFIRST today.

Input: A written description, plus up to 4 optional reference photos
You get: One new image
Formats: 1:1 square, 16:9 wide or 9:16 vertical
References: Products, people, places or a style, uploaded or picked from References
Where it goes: Your Data Room, ready to open on the board of a session

Questions people ask.

About Create image and how it fits the rest of DFIRST.

How do I make on-brand images with an AI image generator?

Describe the picture in the box above and add reference photos of what must stay true: your product, a face, a colour world. DFIRST sends the words and the photos to the image model together, so the result follows your references rather than a generic look.

Can I use a reference image?

Yes, up to four. Upload them or pick them from References. Use one for the product, one for a person and one for the light or style, and say in your description what each one is for.

What should I write in the description?

Say what is in the frame, where it is, the light and the mood, in plain sentences. Add the framing too: close-up, wide shot or flat lay. Short, concrete descriptions work better than long lists of adjectives.

Which formats can I choose?

Square 1:1, wide 16:9 and vertical 9:16. Each run makes one picture. To see the same idea in another shape, run it again with a different format or change the format of the card on the board.

Can I switch to another image model?

Yes, on the board. Open the result in a session and the image card lets you pick another model, format and quality, then draw it again. The first take stays, so you can compare the two.

Is anything posted for me?

No. The image is saved in your workspace and stays there until you download it, open it on a board or use it in another tool. DFIRST makes drafts: nothing is published, posted or sent anywhere without you.

All 55 tools