Kling 3.0 — multi-shot video with sound from text or frames
Kling 3.0 is Kuaishou's video model: describe a shot or give it a first and last frame, and get a 3–15 second clip in std, pro or 4K, with native sound if you want it, from $0.08064 per second.
- Text to video
- First and last frame
- Native sound
- Up to 5 shots
- Up to 4K
- 3–15 seconds
Try Kling 3.0 before you write any code
Run Kling 3.0 with the exact request your code will send. Every parameter the API takes is here.
kling-3-0from $0.08064/s
Describe the shot, add a first or last frame if you have one, pick the mode and length, and submit once. The Kling 3.0 clip arrives with the exact amount it cost.
Subject, place, camera move and light. Up to 2,500 characters. Needed unless multi-shot is on.
Files are uploaded and sent as URLs.
Optional. The first image opens the clip; a second one is where it ends. Without images, the clip is made from text.
std costs least, pro is sharper, 4K is the highest quality.
3–15 seconds. In multi-shot mode the shots set the length instead.
16:9 for YouTube, 9:16 for Reels and TikTok, 1:1 for feeds.
On generates sound with the picture and costs more per second.
On builds the clip from the shots below instead of one prompt.
A JSON list of up to 5 shots, each with a prompt and 1–12 seconds. The shots add up to 3–15 seconds.
A JSON list of up to 3 subjects to keep consistent: a name, a description and 2–4 image URLs. Mention each by name in the prompt.
Estimated at current rates from the mode, the sound setting and the length. In multi-shot mode the shot durations set the length.
Try one of these prompts
Generated with Kling 3.0 on SeedRouter
Kling 3.0 pricing
Kling 3.0 is billed per second of video at a rate set by the mode and by whether sound is on. Every rate below is the one that bills your account.
- Kling 3.0
- $0.5376/clip
- Kling 3.0
- 248seconds
All Kling 3.0 prices come from the rates that bill your account. Failed requests are not charged.
| Category | Price · 1 second |
|---|---|
| std | $0.08064 |
| std & sound | $0.12096 |
| pro | $0.10752 |
| pro & sound | $0.16128 |
| 4K | $0.514272 |
| 4K & sound | $0.514272 |
What is Kling 3.0?
Kling 3.0 is the third generation of Kuaishou's Kling video model. One Kling 3.0 call turns a prompt, or a first and last frame, into a 3–15 second clip, and can add native sound in the same render.
Kling 3.0 can also cut a clip into up to five shots, each with its own prompt and length, and keep up to three subjects consistent from reference images. On SeedRouter you send one request, get a task ID, and collect the MP4 when it is ready.
In our tests, a 3-second std Kling 3.0 clip came back as MP4 (H.264) at 1280 × 720, and a 5-second pro clip with sound at 1920 × 1080 with its audio track, each in about two to three minutes.
Two ways to use Kling 3.0
Call the API from your backend, or hand the job to a coding agent.
Call the API
Send the Kling 3.0 request from your server, keep the task ID, and read the finished clip from the task.
- 1Create an API key
- 2POST the request with your prompt
- 3Poll the task until the video is ready
Let an agent do it
Copy a ready prompt into Claude Code, Codex or Cursor. It asks before spending and downloads the result.
- 1Copy the prompt
- 2Paste it into your agent and describe the shot
- 3Approve the cost, then get the video
What Kling 3.0 is good at
Kling 3.0 is at its best when a scene needs more than one camera angle, or when it has to start and end on your own frames.
A scene, not just a shot
Write up to five shots with their own prompts and lengths. Kling 3.0 cuts them into one clip.
Start and end where you want
Give Kling 3.0 a first frame, a last frame or both, and the clip moves between them.
Sound when you need it
Turn sound on and the clip arrives with it. Leave it off for silent b-roll at a lower rate.
The same face in every shot
Add up to three elements from 2–4 photos each, and Kling 3.0 keeps that person or product consistent.
Draft cheap, finish in 4K
Draft in std, then render the final take in pro or 4K. Same request, one field changes.
Up to 15 seconds
One Kling 3.0 call covers a whole short scene, from 3 to 15 seconds.
What teams use Kling 3.0 for
Product spots
Keep a real product consistent as an element and cut a three-shot spot around it.
Social clips
Kling 3.0 clips in vertical 9:16 for Reels and TikTok or wide 16:9 for YouTube.
Storyboard to motion
Turn a key frame and an end frame into a Kling 3.0 shot that joins them.
Recurring characters
Use the same element across many clips so a mascot or host looks the same every time.
B-roll at scale
Generate silent std clips in batches for edits, explainers and presentations.
Trailers and teasers
Cut 4K Kling 3.0 shots into a teaser, with sound rendered in.
Start with Kling 3.0 in four steps
- 01
Create a key
Sign in, top up any amount, and create an API key. Credits never expire.
- 02
Send one request
POST your prompt to /v1/videos/generations with model kling-3-0, plus frames, mode and length if you need them.
- 03
Poll the task
Keep the Kling 3.0 task ID and check it every 10–20 seconds until the status is completed.
- 04
Download the video
Save the hosted Kling 3.0 URL from the task to your own storage.
Kling 3.0 at a glance
The facts that decide whether Kling 3.0 fits the job.
- Quality
- std, pro, 4K
- Clip length
- 3–15 seconds
- Inputs
- Text, first and last frame, element images
- Audio
- Optional native sound
- Multi-shot
- Up to 5 shots per clip
- Billing
- Per second, by mode and sound
Need reference videos or audio? See MiniMax H3. Need clips longer than 15 seconds? See Seedance 2.5.
What to know before you start
Shots add up to 3–15 seconds
In multi-shot mode each Kling 3.0 shot runs 1–12 seconds and all of them together 3–15.
Images are sent as URLs
Send Kling 3.0 frames and element photos as public HTTP(S) links, not base64. The Playground uploads frames for you.
Results arrive by task
A Kling 3.0 request returns a task ID first. Poll it until the video is ready.
Every element needs a description
Each element takes a name, a short description and 2–4 images of the subject.
Why run Kling 3.0 on SeedRouter
Pay per second of video
Kling 3.0 is billed on the seconds you ask for, at the rates in the table above. No subscription, no minimum.
Failed requests are free
If a Kling 3.0 task ends failed, nothing is charged. You only pay for videos that are delivered.
Credits never expire
Top up once and use the balance whenever you need it.
Hosted results
Finished Kling 3.0 clips come back as URLs on our CDN. Keep a copy of anything you need long term.
One balance for every model
The same top-up pays for video, image and text models in the catalogue.
Other models on SeedRouter
Same balance, same task-based workflow.
Kling 3.0 guides
Practical guides to pricing, prompting and comparing the model.
Use the Kling 3.0 API step by step: create a key, send a video task, poll it for the video URL, start from a first and last frame, and build multi-shot clips.
Kling 3.0 API pricing per second at 720p, 1080p and 4K, with and without native audio, what a 5, 10 or 15 second clip costs, and how multi-shot is billed.
How to prompt Kling 3.0: the five parts of a strong video prompt, camera language that works, multi-shot scenes, dialogue with sound, and consistent characters.
Kling 3.0 Turbo vs Kling 3.0: what each model can do, the price per second with and without sound, what our tests showed, and which one to pick.
Kling 3.0 FAQ
What is Kling 3.0?+
Kling 3.0 is Kuaishou's third-generation video model. It makes one clip of 3 to 15 seconds from a text prompt or from a first and last frame, in std, pro or 4K, with optional native sound. A clip can hold up to five shots and keep up to three subjects consistent.
How much does Kling 3.0 cost?+
A 5-second pro clip without sound costs $0.5376. Kling 3.0 is billed per second at a rate set by the mode and by whether sound is on, starting at $0.08064 per second in std. The pricing table on this page shows every current rate.
Is Kling 3.0 free?+
Not on SeedRouter. Kling 3.0 is paid per second of video, and new accounts get a small credit to try it. There is no subscription: you top up any amount, credits never expire, and failed requests are not charged.
How do I use the Kling 3.0 API?+
Create a SeedRouter key and POST your prompt to /v1/videos/generations with model kling-3-0, plus mode, duration and aspect_ratio if you want other than the defaults. You get a task ID back; poll /v1/tasks/{id} until the video is ready.
How does multi-shot work in Kling 3.0?+
Set multi_shots to true and list up to five shots in multi_prompt, each with its own prompt and a length of 1 to 12 seconds. The shots must add up to 3–15 seconds, and that sum is the clip length you are billed for.
Does Kling 3.0 generate sound?+
Yes, when you ask for it. Set sound to true and Kling 3.0 renders sound together with the picture. Sound costs more per second, so leave it off for silent clips.
Can Kling 3.0 animate my own image?+
Yes. Pass one image URL in image_urls and Kling 3.0 starts the clip from it; pass two and the clip ends on the second. Images go in as public URLs; the Playground uploads them for you.
Does Kling 3.0 support 4K?+
Yes. Set mode to 4K for the highest quality. 4K costs more per second than std and pro, so many teams draft in std and render the final take in 4K.
Generate your first Kling 3.0 clip
Try Kling 3.0 in the Playground, then move the same request into your code.