π Want to recreate viral AI videos without guessing the prompts?...

Gemini can help you analyze reference videos, break down their visual elements, and uncover detailed prompts to recreate a similar style.
No expensive tools needed. You can get started for FREE!
Hereβs a simple 6-step workflow to turn any reference video into an AI-ready prompt. π§΅π
Start by choosing an AI-generated video that catches your attention. Save it as a visual reference for your recreation.
Before moving forward, break down its key elements:
πΈ Shot composition and camera perspective
π Character behavior, gestures, and emotions
π‘ Lighting setup and visual atmosphere
π¨ Color palette and cinematic effects
β±οΈ Scene pacing, cuts, and transitions
The goal is to understand what makes the video visually engaging before recreating it.
Now upload your saved reference video to Gemini and give it this instruction:
Copy this prompt:
"Act as a professional AI filmmaking director and video prompt specialist.
Analyze the uploaded video in detail, breaking it down into individual scenes in chronological order.
Pay close attention to:
πΉ Subject details, clothing, appearance and positioning
πΉ Background, environment and surrounding objects
πΉ Character movements, gestures and interactions
πΉ Camera angles, zooms, pans and tracking shots
πΉ Lighting, shadows, colors and cinematic atmosphere
πΉ Scene cuts, transitions and overall timing
πΉ Facial expressions, emotions and body language
After analyzing the footage, create a highly detailed AI video generation prompt optimized for Seedance 2.5.
Make the prompt clear, structured and ready to copy and paste.
Describe each scene accurately, including camera direction, subject movement, framing, composition and lighting.
Focus on matching the original video's visual storytelling as closely as possible.
Avoid generic descriptions or assumptions. If something isn't clearly visible, mention the uncertainty instead of making it up."
Want to see yourself starring in that AI video? Upload a clear portrait photo of yourself to Gemini and use this instruction:
Copy this prompt:
"Use my uploaded photo as the identity reference for the main character in the analyzed video.
Replace the original character with the person shown in my photo while maintaining their identity and recognizable appearance.
Pay special attention to:
β¨ Preserve facial structure and distinctive features
π Match hairstyle and hair details
π¨ Maintain natural skin tone and texture
π¬ Follow the original character's movements and expressions
πΈ Keep the same camera angles, framing and composition
π‘ Retain the original lighting, environment and transitions
Make the character look consistent across every scene, with realistic movements and natural integration into the footage.
Do not alter the original storyline or visual atmosphere."
Before generating your video, give Gemini one more task: refine the prompt to make the final result closer to your reference.
Copy this instruction:
"Act as a senior AI video prompt editor.
Review the current video prompt and find the three biggest areas where more detail or precision could improve the final recreation.
For each one:
π Identify what's missing or unclear.
π οΈ Explain how to make it more precise.
π¬ Update the prompt with the necessary improvements.
Then rewrite the complete video prompt, keeping the original scene order, character actions, camera movements and visual style intact.
Make the final version detailed, consistent and ready for AI video generation.
Do not remove important details or introduce unrelated elements."
Now it's time to turn your refined prompt into an actual AI video!
Take your final prompt and paste it into Seedance 2.5 or any compatible AI video generator.
Once the video is generated, compare it with your reference and check:
π€ Character identity and consistency
π Natural movements and realistic actions
π· Camera angles and positioning
π‘ Lighting, shadows and visual quality
β±οΈ Scene order, transitions and timing
Not satisfied with the result?
Go back to Gemini, identify the parts that need fixing, adjust those specific instructions and generate another version.
Keep refining until you get the look you're aiming for.
Don't let your best prompts get lost! Save the versions that deliver great results and turn this method into a repeatable creative process.
You can reuse the workflow for:
π Different characters and identities
π New locations and cinematic environments
ποΈ Product showcases and promotional videos
π¨ Fresh storytelling and creative concepts
π¬ Different AI video styles and experiments
The idea is simple: stop starting from zero every time.
Let Gemini handle the visual breakdown, refine your instructions, and use those prompts to create your next AI video.
One workflow. Endless creative possibilities! π
π Follow @youraigirl24 for more mind-blowing AI Tools and Tricks. x.com/youraigirl24
