Instagram Notification Crush: Photo-to-Video Prompt
Turn your photo into a four-second Instagram notification hook: grab, crumple and throw the banner. Copy the image prompt and exact video prompt.
Reference video; not a personalized output from this prompt.
Read the full prompt
INSTAGRAM NOTIFICATION CRUSH — PHOTO + 4-SECOND VIDEO HOOK Use your own or authorized portrait. Copy PART 1 into ChatGPT with your photo. Download the finished notification image, then upload it to Google Flow and paste PART 2 only. Do not submit both parts together. Change rahulgaur3d and this video hook consistently in both prompts when using your own account name and message. The video prompt requests four seconds and 4K; choose the duration and export resolution actually offered by your tool. The supplied example is four seconds at 1080 × 1920. PART 1 — IMAGE PROMPT FOR CHATGPT Create a photorealistic vertical 9:16 image using my attached portrait as the identity reference. This will be the first frame of a short video in which I grab an Instagram-style notification banner above my head, crumple it like paper and throw the ball forward. CUSTOM TEXT DISPLAY NAME: rahulgaur3d NOTIFICATION MESSAGE: this video hook Use these exact words unless I replace them with my own name and message. Keep the same spelling in the image and video prompts. PERSON AND COMPOSITION Show exactly one person from the supplied portrait, standing in a modern studio. Preserve the recognizable facial features, skin tone, apparent age, hair, facial hair and visible clothing from the photo. Complete unseen clothing with plain coordinated casual garments. Frame from the head to the upper thighs. Both forearms and hands must be visible, relaxed near the torso at the beginning, ready to reach upward. Use realistic anatomy, separate fingers and natural proportions. Keep comfortable space above the head and on both sides for the reach and throw. Do not copy the face of a creator from a tutorial example. NOTIFICATION BANNER Place one horizontal rounded Instagram notification banner above the person's head, within natural reach and entirely inside the image with clear margins. Give the banner a light gray translucent surface and subtle physical thickness. On the left, show the Instagram gradient camera logo. To its right, show the exact display name in bold dark text and the notification message on the next line in smaller dark text. Keep all text legible. The banner is a fictional visual prop, ready to be gripped at its left and right edges. Leave a gap between the banner and the person's head. Do not place it behind the person or already in their hands. STUDIO AND LIGHT Use a clean contemporary studio with restrained neutral gray surfaces, soft practical ceiling lights and a gently blurred background. Light the face naturally and evenly. One fixed eye-level camera, natural perspective, detailed skin and realistic fabric. Keep the person and banner sharp and the studio free of distracting text or other people. AVOID Extra notifications, duplicate people, extra fingers, warped hands, hands covering the face, cropped banner edges, misspelled text, a verification badge, floating unconnected letters, phone frames, captions, watermarks, a ball already present, cartoon rendering or extreme beauty filters. Deliver one still image. Check the face, fingers, name and message before animating. END OF PART 1 PART 2 — EXACT 4-SECOND VIDEO PROMPT A 4-second realistic 4k video. The man from the reference image is standing in a modern studio with a floating Instagram notification banner above his head showing the Instagram logo, bold text "rahulgaur3d", and message "this video hook". He raises both hands, grabs the floating notification banner, crumples it like a real piece of paper into a tight ball, and throws it forward out of the frame while smiling. END OF PART 2
Make a floating Instagram-style notification feel like a physical object. With this notification crush AI video prompt, the person reaches up, grabs the card, squeezes it into a paper ball and tosses it away. The effect works as a playful visual opening for a tutorial or creator Reel.
This guide includes an image prompt and the exact short video prompt supplied for the new example. Use Part 1 to prepare a standing studio photo with a notification banner, then paste Part 2 into Google Flow with that finished image. The supplied preview is a four-second, 1080 × 1920 AI-generated example. Results with another portrait can differ.
Exact four-second video prompt
The copyable prompt above includes the following video section exactly as supplied:
A 4-second realistic 4k video. The man from the reference image is standing in a modern studio with a floating Instagram notification banner above his head showing the Instagram logo, bold text "rahulgaur3d", and message "this video hook". He raises both hands, grabs the floating notification banner, crumples it like a real piece of paper into a tight ball, and throws it forward out of the frame while smiling.
Keep your own portrait as the identity reference. Replace the sample account name and message if needed. The phrase 4k is a requested quality target, not proof of the generated file's resolution. The supplied preview is 1080 × 1920, not 4K. Check your actual export settings and downloaded file.
What the notification crush effect does
The visual begins as a familiar floating social notification above a person. The hands make contact with its edges, pull it down and turn the flat card into flexible paper. The paper folds into one ball before a small throw removes it from the scene. The surprising moment is the transition from a screen-like graphic to a tangible prop.
The new example shows a man in a striped shirt standing in a modern studio. The banner reads rahulgaur3d and this video hook. He grabs it, crumples it and throws it forward. Use your own portrait, account name and message for a personalized version. The notification is a fictional graphic created for the scene.
Prepare a useful portrait
Choose a clear photo of one person with an unobstructed face and natural lighting. A standing waist-up or upper-thigh portrait showing both arms is helpful. The first frame needs enough headroom for a card and enough room for both hands to reach it. A tight face crop provides less information about the person's body and clothing.
Use your own photo or an image you are authorized to use. Keep the reference simple, without extra people, text overlays or heavy filters. If a full outfit is not visible, the prompt completes only the missing areas with plain casual clothing instead of pretending the unseen details are known.
Customize the name and message
In Part 1, edit DISPLAY NAME and NOTIFICATION MESSAGE before submitting. The included example is rahulgaur3d with the message this video hook. Replace both with the wording you want displayed, and use the same text in Part 2. A short message of roughly four to eight words is easier to fit than a paragraph.
Keep the name and message on separate lines. Avoid requesting a verification badge or implying that the message came from a real person who did not send it. The notification is a visual storytelling prop. Check the spelling in the finished photo before moving to video.
Create the starting image in ChatGPT
Open ChatGPT, sign in if needed, attach your portrait and paste only Part 1: Image Prompt. Ask it to create the image. Use the image creation option if your interface requires selecting it. OpenAI's Images in ChatGPT guide explains that an existing image can be uploaded and edited by describing the desired change.
Review the resulting composition. The person should match the portrait, both hands should be visible and the card should sit above the head within reach. Download the image once the card text and face look correct. Access and usage limits depend on your account; this guide does not promise unlimited generation.
Check the photo before animating
The starting image carries most of the visual information into the animation. Inspect the full name, message, icon, card edges, hands and face. A misspelling or an extra finger in the still image can persist in the video.
If the text needs repair, request a targeted edit that changes only the lettering and keeps the composition intact. If the card is too close to the top edge, move it downward while retaining a gap above the head. Do not place it behind the person or already inside the hands; that makes the reach-and-grab sequence less clear.
Set up the image in Google Flow
Open a Flow project and select a video workflow that accepts a starting image. Add the finished notification photo as the start frame or corresponding image input. Choose portrait 9:16 and a four-second duration if your selected tool offers it. Otherwise use an available duration and trim the finished action during editing. Upload the generated photo, not the original portrait without the notification.
Google's Flow video creation instructions describe creating video with image inputs and adding a start frame. Labels and supported controls vary by model and interface. If a rights notice appears, review it and confirm only when you have the necessary permission. Check the selected format before submitting.
Paste the motion prompt
Paste only Part 2: Exact 4-Second Video Prompt into Flow. It describes one continuous action: reach, grip, pull down, crumple, toss and relax. The image supplies the identity and scene, while the text supplies the movement. Do not paste both prompts together into the video box or ask Flow to recreate a different person from the tutorial reference.
The short prompt requests a fast reach, grab, paper crumple and forward throw while smiling. Inspect the result for visible hand contact and progressive folds. A fixed camera helps the audience understand the effect. The card should move with the hands after the grip, and the printed text should bend with the paper. It should not remain as a second floating overlay.
Review the hand and paper motion
Generate when your inputs are ready, then inspect the result through the whole action. Watch for fingers passing through the card, two balls appearing, text floating after the paper has folded or the notification returning at the ending. The throw should follow the hand motion and leave the frame in a consistent direction.
The prompt guides the model but cannot guarantee exact identity, spelling or physically perfect fingers. The preview is the supplied example clip, not a new test using every visitor's photo. Its appearance and hand motion are illustrative; matching a different uploaded portrait has not been tested. If the motion is wrong, revise the specific problem and regenerate with the same start frame instead of adding unrelated style instructions.
Edit and download the final clip
Download the finished clip using the controls available in your Flow interface. Add your own music during editing. Quiet paper crinkles can emphasize the squeeze, while a subtle click can support a tutorial step. Keep the voiceover separate from any reference speaker.
For a short tutorial Reel, a brief music-led preview can establish the effect before the upload and prompt steps. Use readable captions, a clear pointer and a few deliberate transitions. A cover featuring your own or an original fictional character with a short title makes the topic easier to recognize. Select that cover manually when the publishing platform offers the option.
Frequently asked questions
Do I need two prompts? Part 1 prepares a notification image and Part 2 animates it. If you already have a suitable starting image, you can use Part 2 directly. Do not paste both parts together into the video tool.
Can I change the room or clothing? Yes, but keeping a simple setting and consistent outfit makes it easier to review the motion. The portrait should remain the identity reference.
Why does the notification disappear instead of folding? Ask for continuous finger contact, progressive paper creases and one visible ball. Remove instructions for magic, explosions or particle effects.
Will this guarantee a viral Reel? No. The prompt creates a visual concept; audience response depends on the hook, execution, audience and publishing choices. Explore more ideas in the AI Prompt King library.