how to use
Open ChatGPT, attach the photo (or type the topic, for the content ones) and type the command. Tap a command to copy it.
They're not an official ChatGPT feature: they're shortcuts the AI understands by name. That's why they also work in Gemini and other image AIs.
If the result doesn't come out the way you pictured, add what you want in one sentence. E.g.: /miniature, keep my face the same.
image
Photos of people or places
/fisheyephoto with a fisheye lens/bigheadcamera up close, face in focus/giantobjectwhat's in the photo becomes giant in the landscape/underwaterthe place ends up underwater/miniaturethe scene turns into a toy model/legoturns into a Lego set, with box and manual/timemachineredoes the photo as if it were from another era/hologramturns into a sci-fi hologram/polaroidturns into an instant camera photo/filmanalog film look, with grain/noirdramatic black and white, old cinema style/goldenhourgolden late afternoon light/editorialfashion magazine photo/streetstylestreet fashion photoimage
Improve your photo
/enhanceimproves quality without changing who's in it/headshotprofessional headshot, company profile style/linkedinphoto ready for your LinkedIn profile/studio-bgswaps the background for a clean studio/blur-bgblurs the background so you stand out/restorerecovers an old or damaged photo/colorizecolors a black and white photoimage
Product photos
/explodedviewthe product comes apart into floating pieces/cutawaycuts it in half and shows what's inside/3dbillboardthe product on a giant billboard in the city/showcasea simple photo becomes a studio shoot/metaadsad artwork ready for social media/lifestyleshotthe product in a real-life scene/floatingproductthe product floating, clean and minimal/gravitydefyingthe product in an impossible scene, defying gravity/ingredientexplosionthe ingredients burst around the product/macroshotextreme close-up on the details/flatlaytop-down, organized shot, online store style/luxurycampaignlooks like a luxury brand campaign/ugcmodela person using the product, creator post stylecontent
Ideas and content
/infographicturns a topic into an infographic/mindmaphand-drawn mind map/storyboardvisual script in 6 panels/socialcarouselInstagram carousel idea/thumbnaileye-catching video cover/posterone-page poster/ebookcovercover for an ebook or free resource/wireframerough draft of a site or app screen/isometricangled 3D illustration, digital model style/blueprintturns into a blue technical drawingvideo
Video prompts for Magnific
In Magnific, attach the image, paste the prompt and generate.
A hand pinches the top of the printed portrait photo held in hands and peels it downward off the frame, revealing the city street scene underneath, realistic paper peel with slight bend, static camera, same lighting
Smooth 360 degree orbit shot around the person, starting from the front, the camera circles all the way around at eye level and returns to the front, the person stays still and natural, consistent face and outfit, realistic
The woman walks naturally along the sidewalk holding her latte. Suddenly she trips on the pavement and stumbles forward, and the coffee flies out of the cup. The moment of the splash turns into extreme slow motion, almost frozen: the coffee splash and droplets hang suspended in the air while the camera keeps moving slowly toward her side. Realistic, cinematic, normal lens, same woman and same street.
Static locked camera, no zoom, no pan. The thick melted cheese keeps flowing and dripping continuously down from the burger at the top of the frame, running down over the black Instagram UI bar with the icons and caption, pooling and splattering into the cartoon character's open mouth at the bottom. Realistic viscous cheese physics, slow continuous drip, everything else in the frame stays completely still like two stacked photos in a feed, only the cheese moves.
Tip: in the video prompts, swap "latte", "burger" or "woman" for whatever is in your photo.