Two Ways to Turn One AI Image Into a Video (And When to Use Each)
One method for freedom. One for seamless scene transitions. Both inside.
TLDR: Today I will show you how I made these two AI videos — two different ways you can do it and which option you want to choose for what scenario. You will also get the prompts of the image & video that I used for them at the end.
Use the Ultimate Prompt Creator (free, no signup):
Want access to the UPC Image-to-Video GPT?
Getting started
So let’s get started then. First of all, I made the image. (Prompt at the bottom 👇) I am not going to explain now how I made that, but if you want to know how I make my AI images, go check this out (more content regarding this will follow in the future): Hogwarts Common Rooms Reimagined With AI Prompts
Now this is a really awesome image that I genuinely really love. It is almost Transformers meeting ethereal fantasy and crystals. 😋🔥
But anyway, from here we have two different paths we can choose to use. Which one you will choose will always depend on the specific use case you need it for, but I will cover both options:
1: Just a starting reference: This is great if you just want to bring it to life, or use it for something else that doesn’t have a next scene that will start at the exact same frame that this video should end on. So let’s say, for example, you use this AI video as the first scene, and then the next scene switches to something completely different. Then this is great, because in my experience the creativity is better!
2: Beginning to end frame: This is for if you have multiple scenes that need to follow each other up at the exact same scene. So let’s say the first scene is floating over a city, and then the second scene continues exactly where the first scene left off. Then this is great because you can achieve high accuracy with this!
So having said that, let’s start with option one 👍💪
Just a starting reference
You open the video prompt GPT, upload the image, and simply tell it what you want it to do. You can keep this as basic or advanced as you like. You do need to think about the difference though. If you keep it basic, the AI will be able to throw in more creativity. If you describe more, it will do as you wish — but if you pin it down too much, it will lose creativity. 🤔
Here is what I entered. And for those paying attention, ignore the typo “wiffing” — it is supposed to be “fitting.” It’s a good thing AI can interpret the worst typos easily. 👀😅😂
Then based on that, it gave me the prompt instantly. (Prompt is at the bottom 👇) Now all that is left to do is to throw it into your desired AI video generator. I personally threw it into Freepik.ai and used the model Kling 3.5. Why? It didn’t consume credits, even though the result is great. 😋
The Result
Beginning to end frame
For this, you will need to upload both the starting frame and the end frame. So here you will need to think about which frame it should end on. Keep in mind that the end frame of this video will be the starting frame of the next video for a frictionless continuation of the scene.
You will need to use an AI image generator that is capable of consistent characters & scenes. For me that is ChatGPT, but there are ofc many other options out there. 👍🤔
So in my case, I just asked ChatGPT to put a sword into its hand and put him in a combat stance, and I got a good image. Then you upload both the beginning and the end frame into the Video Prompt GPT, tell it that it is for beginning–end frame, and tell it what you want. This is what I entered:
Then I got the prompt and put it into Freepik again.
For this, you need to choose a model that also supports beginning and end frames, as not all AI models support that.
The Result
End note
That is it. Literally.
Then with this you can do whatever you want. But… there is one warning I want to add: if you use the free tier of a video generator, then check the rules. For many AI video generators, you are not allowed to earn money with the videos in any way if you are on the free tier. Not all of them — but many.
So if you are planning to use it for brand building or something else, and it leads to income in any way possible, make sure to check that first. You want to look for “commercial use” or something along those lines. 👀
Use the Ultimate Prompt Creator (free, no signup):
Want access to the UPC Image-to-Video GPT?
The prompts:
THE IMAGE PROMPT:
Description: A towering crystal golem stands at the center of a vast jewel-toned crystal chamber, its massive body composed of faceted sapphire, emerald, and amethyst crystal formations glowing from within. Over its luminous gemstone structure lies brutal, spiked armor forged from polished dark steel and radiant gold accents, wrapping around its shoulders, chest, and forearms like ceremonial war plating. The armor is jagged, angular, and menacing, enhancing its already colossal silhouette. A powerful waterfall cascades behind it, refracting prismatic light across its crystal body. Monumental crystal pillars surround the chamber in vibrant jewel tones, while above, the ceiling opens into an endless cosmic expanse of stars and suspended crystal shards. The lighting is cinematic golden twilight — bright, luminous, and radiant — highlighting every razor edge and reflective surface. The creature is breathtakingly beautiful yet undeniably lethal.
Keywords: #CrystalGolem #JewelWarlord #ArmoredTitan #CinematicFantasy #RadiantMenace #EmeraldSapphire #CrystalChamber #Vertical916
Modifier Tags: [low-angle heroic framing] [vertical 9:16 ratio] [monumental scale] [sharp angular armor design] [internal crystal glow] [high contrast lighting] [volumetric mist] [refraction effects] [hyper-detailed textures] [cinematic color grading] [realistic waterfall physics] [environmental depth clarity] [8k detail]
Negative Prompt: dark horror aesthetic, muddy colors, dull lighting, cartoon style, soft rounded armor, childish design, low detail textures, blurry image, distorted anatomy, low resolution, flat lighting
VIDEO PROMPT OF JUST STARTING REFERENCE (I personally never use the negative prompt, but you can if you whish):
**1. Main subject movements**
The towering crystal-armored golem lifts his head slightly, chest expanding as glowing energy pulses beneath the multicolored crystal plates, he leans forward and unleashes a powerful roar toward the camera, mouth opening wide with visible vibration through the crystal jaw, shoulders rising with force, he then raises one heavy foot and slams it down into the ground with controlled impact, body weight shifting downward realistically, armor plates subtly rattling from the force, after the stomp he extends one arm outward as golden energy gathers around his hand, forming swirling light shards that rapidly condense into a massive glowing crystal sword, fingers tightening around the hilt as the blade solidifies, he rotates his torso and lowers into a grounded combat stance, knees bent, shoulders squared, sword angled forward in a ready position
**2. Secondary subject interactions**
Floating crystal shards around him tremble and vibrate during the roar, nearby large crystals resonate with pulsing light synced to his energy surge, smaller fragments lift briefly from the ground during the stomp shockwave, light reflecting dynamically across his armor facets
**3. Environmental effects**
The ground fractures outward from the stomp in radial cracks, chunks of stone lift and settle with gravity-consistent motion, golden energy ripples travel across the floor surface, waterfalls behind him surge slightly from the shockwave, cascading water scattering fine mist
**4. Atmospheric changes**
Golden particles intensify around his body during the roar, glowing dust swirling in circular motion during sword formation, light rays from above flicker brighter at peak power, settling into a steady radiant aura as he holds his combat stance
## 🚫 **Negative Prompt**
No camera cuts, no sudden scene transitions, no new characters appearing, no exaggerated bouncing physics, no slow motion distortion, no unrealistic limb stretching, no floating without force explanation, no cartoon effects, no exaggerated facial deformation, no overly chaotic debris, no disappearing armor pieces, no weapon appearing instantly without energy buildup, no lighting changes that contradict the golden source, no perspective warping, no frame flicker, no blur overload, no smoke covering the subject, no unrealistic speed shifts
VIDEO PROMPT OF BEGINNING-END FRAME (I personally never use the negative prompt, but you can if you whish):
Crystal-armored golem warrior stands tall and motionless at first, chest slowly rising with controlled breath, faint internal light pulsing through multicolored crystal plates, head slightly tilts downward then lifts with rising intensity, fists tighten gradually as golden armor edges catch light, torso rotates subtly to the right while left arm lowers with deliberate force, right hand opens as glowing particles gather and swirl inward, concentrated energy forms between his fingers, a radiant sword begins to materialize from bright crystalline light, blade extending outward in a smooth upward motion, golem steps forward half a pace and powerfully raises the fully formed sword into a dominant battle stance, shoulders squared and posture widened, energy stabilizing around the blade,
Floating crystal shards in the air begin to rotate slowly then accelerate in synchronized orbit around him, nearby tall crystals emit brighter reflections as the sword forms, subtle tremor runs through the ground as he shifts stance,
Golden waterfall behind him flows continuously with slightly intensified brightness during the sword materialization, sparkling particles cascade downward in increased density,
Atmosphere glows warmer and more radiant, ambient light blooms softly around the sword, faint energy waves ripple outward through the air, epic and powerful mood maintained, no camera cuts, smooth transition from still stance to battle-ready pose.
## 🚫 Negative Prompt
No camera cuts, no sudden jumps in position, no unrealistic fast movements, no exaggerated limb distortion, no bouncing motion, no floating body, no additional weapons, no new characters, no background changes, no environment replacement, no physics-defying transformations, no melting armor, no warping geometry, no low detail textures, no flickering light artifacts, no inconsistent crystal colors, no disappearing elements, no time skips, no overly fast sword appearance, no chaotic motion blur.