Kling 3.0 prompts: 3 stills brought to life with image to video
Kling 3.0 Standard prompt on Twistavo, an AI video studio for character swap: replace a person in a real video from 1 to 4 photos; plans from $3.99/month.
Updated
Kling 3.0 image to video starts from a picture you already have and animates it into a short clip. On October 5, 2026 we made three street-photo stills with Soul 2 and gave each one to Kling 3.0 Standard with a short motion prompt: a man stepping out of a café, a woman cycling through a park, a man walking down an avenue. Below are both prompts for each clip, the settings, the credit cost, and where the still and the motion prompt did not agree.
Make your own with Twistavo — 1 free imageSetup
- Model
- Kling 3.0 Standard
- Settings
- Two steps per clip. Step 1: Soul 2 text to image, 16:9, 1080p, one image, prompt enhancement off. Step 2: Kling 3.0 Standard image to video with that image as the start frame, 16:9, sound off, CFG scale 0.5, at the duration listed under each example, multi-shot off. The motion prompt describes only what moves and how the camera moves; the look comes from the still. In Twistavo: open the AI Studio, pick Kling 3.0 Standard, upload the still in the start-frame slot, paste the motion prompt, set that duration and 16:9, then generate. The cost under each example is the clip; the Soul 2 still is priced separately.
- Models used
- Kling 3.0
1. A man steps out of a café and takes a sip
Run record not kept: prompt and settings are the team's notes, not checked against the run.
- Kling 3.0 Standard, image to video
- Start image: the Soul 2 still
- Aspect ratio 16:9
- Duration 5 s
- Sound off
- CFG scale 0.5
54 credits in Twistavo
Prompt
Start image (Soul 2, 16:9, 1080p): Cinematic photo of an adult man in his forties in a plain navy overshirt stepping out of a cafe onto a sunlit terrace holding a cup, medium shot, face clearly visible, warm afternoon light, city street behind, 35mm, absolutely no text, no logos, no prints, no signage Video (Kling 3.0 Standard, image to video): The man steps out of the cafe doorway onto the sunlit terrace holding his cup, takes a sip and looks down the street, slow cinematic camera push-in, warm afternoon light, no text
Tip: Kling kept the face, the shirt and the cup from the still for the whole clip and added the sip. It did not follow "looks down the street": at the end he looks into the lens. The still prompt said "no signage", yet the street behind has red shop signs; they stay blurred, but check the background of a still before you animate it, because the video keeps everything the still has.
2. A woman cycles toward the camera in a park
Run record not kept: prompt and settings are the team's notes, not checked against the run.
- Kling 3.0 Standard, image to video
- Start image: the Soul 2 still
- Aspect ratio 16:9
- Duration 5 s
- Sound off
- CFG scale 0.5
54 credits in Twistavo
Prompt
Start image (Soul 2, 16:9, 1080p): Cinematic photo of an adult woman in her thirties in a plain olive jacket riding a bicycle along a sunny park path lined with trees, medium shot, face clearly visible, relaxed smile, warm morning light, 35mm, absolutely no text, no logos, no prints Video (Kling 3.0 Standard, image to video): The woman rides her bicycle toward the camera along the sunny park path, smiling, natural pedalling, the camera tracks backward smoothly, trees and lawn around, warm morning light, no text
Tip: The still prompt asked for an olive jacket; Soul 2 made a dark brown leather jacket and added glasses. Kling then kept that jacket and those glasses in every frame: image to video copies the still, not the words of the first prompt. Fix the still first, then animate it.
3. A man walks down a sunny avenue
Run record not kept: prompt and settings are the team's notes, not checked against the run.
- Kling 3.0 Standard, image to video
- Start image: the Soul 2 still
- Aspect ratio 16:9
- Duration 5 s
- Sound off
- CFG scale 0.5
54 credits in Twistavo
Prompt
Start image (Soul 2, 16:9, 1080p): Cinematic street photo of an adult man in his thirties wearing a plain white jacket over a plain grey t-shirt with no print, dark jeans, riding a skateboard down a wide sunny city avenue, medium-wide shot, full body and face clearly visible, relaxed smile, tall buildings and trees behind, warm natural daylight, 35mm, absolutely no text, no logos, no prints Video (Kling 3.0 Standard, image to video): The man walks confidently toward the camera down the sunny avenue, smiling, natural arm swing, the camera tracks backward smoothly in front of him, cars and trees in the background, warm daylight, no text
Tip: The still prompt asked for a skateboard and Soul 2 drew the man striding with no board. The motion prompt was written for what the still shows ("walks confidently toward the camera") and the clip works. Write the motion prompt after you look at the still, never from the first prompt.
Make it in Twistavo
- Open the AI Studio in Twistavo, pick Soul 2 and make the still with the start-image prompt, at 16:9.
- Look at the still before you write the motion prompt: Kling animates what the still shows, not what the first prompt asked for.
- Pick Kling 3.0 Standard and upload the still in the start-frame slot.
- Paste a motion prompt with one action and one camera move, set 16:9, 5 s, sound off and CFG scale 0.5.
- Check the credit quote and generate.
Variations to try
- A longer clip with a second action, such as "then sits down at a table".
- Sound on for street ambience: all three clips here were made with sound off.
- Kling 3.0 Pro instead of Standard, for finer detail at a higher price.
- Fix the still first when it differs from your idea (the olive jacket that came out brown), then animate the corrected still.
Notes
What works with Kling 3.0 image to video: one action plus one camera move per clip, in a single sentence; the subject from the still keeps the same face and clothes for the whole clip. Variations to try: a longer clip with a second action ("then sits down at a table"), sound on for street ambience, or Kling 3.0 Pro for finer detail at a higher price. All three stills and clips were generated with AI on October 5, 2026; the people in them are not real. Results change from run to run.