Kling 3.0 Guide: Start Frames, Prompts, Sound and Credits
How to use Kling 3.0 in the Spicy Studio Video Lab: start and end frames, 5 or 10 second clips, Standard vs Pro, optional sound and what each clip costs.
What Kling 3.0 Does in the Video Lab
Kling 3.0 animates a picture you already have. You upload the opening frame, optionally a closing frame, and describe what should move; the model turns that into a 5- or 10-second clip. It runs in the Spicy Studio Video Lab, and this link opens it with Kling 3.0 selected.
Inputs: a start frame (required; JPEG or PNG, up to 10 MB, at least 300 px), plus an optional end frame and prompt.
Output: 5 or 10 seconds in 16:9, 9:16 or 1:1, at 720p on Standard or 1080p on Pro.
Sound: off by default; switching it on multiplies the price by 1.5.
It suits shots whose look is already settled, such as a finished product photo, an approved character design or a storyboard panel.
What a Clip Costs
You pay per second of output.
Standard (720p): 12 credits a second: 60 for 5 seconds, 120 for 10, or 90 and 180 with sound.
Pro (1080p): 16 credits a second: 80 for 5 seconds, 160 for 10, or 120 and 240 with sound.
On Spicy Basic (€10 for 600 credits a month), a silent 5-second Standard clip works out to about €1. The 25 free credits on a new account won't reach the 60-credit minimum, but they pay for a start frame: Nano Banana 2.1 makes one for 12 credits at 2K. A €2 Spark top-up (90 credits that never expire) then covers a first Standard test. See pricing.
Get the Start Frame Right First
Kling 3.0 moves what's in the start frame; it doesn't redesign it. Whatever is soft, badly cropped or wrong in the still stays wrong for the whole clip.
Crop to the ratio you'll pick (9:16, 16:9 or 1:1), so nothing important sits at the edge, and leave space for any motion, such as a raised arm or a camera pulling back.
Convert WebP files to PNG or JPEG before uploading.
Remove distractions. A stray cup or sign appears in every frame; the Eraser paints it out for 8 credits.
No still yet? Generate one in the Image Lab, or follow the Playbook's first-frame workflow, free with an account.
Prompt the Motion, Not the Picture
The model already sees the image, so the prompt only needs what changes: what moves, what the camera does (static, slow push-in, orbit, handheld) and, with sound on, what we hear. Keep 5 seconds to one main action; 10 seconds has room for two or three, in order.
Product shot (Standard, 5 s, 1:1): “The camera circles slowly around the leather backpack on the oak bench. Late-afternoon light slides across the brass buckle. The bag stays still.”
Portrait (Pro, 10 s, 9:16): “The man in the denim jacket checks his watch, looks up toward the street and breaks into a grin. A light breeze moves his hair. Handheld camera, gentle drift.”
With sound on, end with what we should hear: “Sound: rain tapping on the café awning, distant traffic.”
Choose Where the Shot Ends With an End Frame
The optional end frame decides where the shot finishes: you choose the last picture, and the model works out the motion that gets there.
Transitions: a café terrace at noon as the start frame, the same terrace at dusk as the end frame, and a prompt like “Locked-off camera. Daylight fades to dusk, string lights flicker on, people come and go.”
Reveals: start on a closed gift box, end on the open box with the product inside.
Loops: try the same image as both frames, then check the join before posting.
Keep both frames the same shape and framing; the bigger the jump between them, the more the model invents.
Standard or Pro, Silent or With Sound
Draft on Standard with sound off (60 credits for 5 seconds), then render the keeper the way you'll publish it. A re-render is a new generation, so it won't match the draft frame for frame.
Go Pro when the clip will be watched large or cropped later; it renders 1080p instead of 720p. Video Upscale can take a finished clip to 2K or 4K for 28 or 56 credits.
Turn sound on for ambience and effects made with the picture. If a character must say particular words, Lip Sync uses your own recording instead: it turns a face photo and an MP3 into a talking clip, from 10 credits a second.
When a Different Model Fits Better
More than 10 seconds, or several beats with their own prompts: Kling Omni O3 runs 3–15 seconds with optional prompt segments (Kling Omni O3 guide).
An idea but no image: Seedance 2.5 and Veo 3.1 can start from text.
Moves copied from a real video: Kling 3.0 Motion Control in Character Swap (motion transfer guide).
A still that acts out a small story, or cheaper drafts: the Playbook's Animate workflow recommends Seedance 2.0 or 2.5 over the Kling models for that job, and Seedance 2.0 costs 35 credits for 5 seconds at 480p.
All three models on inputs, length and credits: Seedance 2.0 vs Kling Omni O3 vs Kling 3.0.
Frequently Asked Questions
How much does Kling 3.0 cost?
+
On Spicy Studio, 12 credits a second on Standard (720p) and 16 on Pro (1080p), or 18 and 24 with sound. A 5-second clip costs 60 to 120 credits.
How long can Kling 3.0 videos be?
+
5 or 10 seconds in the Spicy Studio Video Lab. As of October 2026, Kling's own site says Kling 3.0 runs up to 15 seconds there; for that length in Spicy Studio, use Kling Omni O3.
Can Kling 3.0 make a video from text only?
+
Not in Spicy Studio: a start frame is required. Make one in the Image Lab, or use Seedance 2.5, Seedance 2.0 or Veo 3.1, which work from text alone.
Is Kling 3.0 the newest Kling model?
+
No. Kling introduced Kling 4.0 on its blog on September 30, 2026. Spicy Studio's Video Lab offers Kling 3.0 and Kling Omni O3.
Want to run this yourself? Every new account starts with free credits.
Open Kling 3.0 in the Video Lab