Seedance 2.5 is coming soonLearn more
Kling AI family

Kling AI Family

Kling AI is a family of video generation models built by Kuaishou. Use this hub to compare Kling 3.0, Kling 3.0 Turbo, Kling O1, Kling O3, and Motion Control, then jump to the right pricing guide, API tutorial, or model page on EvoLink.

Compare the Kling models

Each Kling model targets a different production need. Use this table to decide which route should own your workflow on EvoLink.

ModelBest forModesEditingDurationEntry pricing
Kling 3.0

Standard

Reliable text-to-video and image-to-video generation for standard production workflows.Text-to-video, image-to-videoNo3-15 seconds$0.075 - 0.150/s
Faster short-form text-to-video and first-frame image-to-video when cost-efficient iteration matters.Text-to-video, first-frame image-to-videoNo3-15 seconds$0.106 - 0.132/s
Kling O1

Unified

Combined generation and editing with consistent characters and scenes across clips.Text, image, video, and subject inputsYes, instruction-based3-20 seconds$0.111/s
Kling O3

Advanced

Full multimodal control including reference-to-video, editing, and all standard modes.Text-to-video, image-to-video, reference-to-video, video editingYes, reference and instruction-based3-15 seconds$0.075 - 0.125/s
Transferring motion patterns from a reference video to a character image.Motion transferNo3-30 seconds$0.113 - 0.151/s

How to decide which Kling model to use

Follow these 4 rules to narrow down your choice.

1

Start with output type

Text-to-video, image-to-video, or reference-to-video — different Kling models support different input modes.

2

Then check editing needs

If you need to refine clips after generation — instruction-based editing or reference control — compare Kling O1 and O3.

3

Then check duration and cost

Longer clips cost more per second. Match your target duration to the model that supports it at the best price.

4

Finally, consider motion transfer

If your workflow requires transferring movement patterns from a reference clip to a character image, Kling 3.0 Motion Control is the dedicated route.

If you already know your task type, find the recommended starting point in the table below.

Choose a Kling model by workflow: generation, editing, multimodal, and motion transfer

Match your primary video task to the right Kling model.

Your taskRecommended startGood fit if...Watch out for
Standard text-to-video or image-to-videoKling 3.0You need reliable clip generation from text prompts or input images without editing complexityNo editing or reference-to-video support
Fast short-form text-to-video or first-frame image-to-videoKling 3.0 TurboYou need faster 720p/1080p generation, 3-15 second clips, and cost-efficient iterationNo reference-to-video or video editing; use a new Kling Open Platform key for direct provider access
Generation + editing in one engineKling O1Character consistency and instruction-based editing matter across multiple clipsSlightly higher cost than Kling 3.0 standard mode
Full multimodal controlKling O3You need reference-to-video, video editing, and all standard modes in one modelMost capable but check per-second cost for your volume
Motion transfer from reference clipKling 3.0 Motion ControlYou want to transfer movement from a reference video to a character imageDedicated to motion transfer — not a general-purpose generation model

Kling API workflows: text-to-video, image-to-video, editing, and motion transfer

See how Kling models fit into real video production pipelines and content workflows.

Text-to-video generation

For marketing clips, social media content, product demos, and explainer videos from text prompts. Start with Kling 3.0 Turbo when speed and cost efficiency matter, or Kling 3.0 when you need the standard route.

Image-to-video animation

For animating product images, illustrations, or photos into short video clips. Start with Kling 3.0 Turbo for first-frame image-to-video, then choose O3 if you also need reference control or editing.

Video editing and refinement

For instruction-based editing, scene refinement, and character-consistent clip iteration. Kling O1 brings generation and editing into one engine. Kling O3 adds reference-to-video on top of editing support.

Motion transfer and animation

For transferring dance moves, gestures, or motion patterns from a reference video to a character image. Kling 3.0 Motion Control is the dedicated route for this workflow, supporting clips up to 30 seconds.

Explore each Kling model

Use this page to compare, then visit individual model pages for pricing details, playground access, and integration guides.

Access Kling models through one EvoLink API

All Kling models are available through a single EvoLink API key. Switch between Kling 3.0, Kling 3.0 Turbo, O1, O3, and Motion Control by changing the model parameter - no separate app integrations needed.

Switch from model="kling-v3-text-to-video" to model="kling-v3-turbo-text-to-video" when you want the Turbo route.
One API key for all Kling models
Async task polling for video generation
Switch models by changing the model parameter
Unified billing with per-second pricing

How to think about Kling video generation cost: quality, duration, and volume

Higher quality increases per-second cost

Advanced modes like Kling O3 reference-to-video and video editing can cost more per second. For bulk short-form testing, start with Kling 3.0 Turbo and reserve O3 for clips that need reference or editing control.

Longer clips multiply total cost

All Kling models use per-second billing. A 15-second clip costs 5x more than a 3-second clip on the same model. Estimate your target duration before choosing a route.

High-volume workflows need cost control

If you generate hundreds of clips per day, per-second cost compounds quickly. Test with Kling 3.0 Turbo for fast text/image generation, then upgrade to O1 or O3 only for clips that need editing, identity, or reference control.

Pricing overview

This page summarizes the pricing shape of the Kling family. Visit each model page for exact live pricing. Kling 3.0 Turbo official list pricing is ¥0.8/s for 720p and ¥1/s for 1080p; EvoLink route pricing is shown on the model page.

Kling 3.0

$0.075 - 0.150/s

per second

Standard generation with per-second billing.

Kling O1

$0.111/s

per second

Unified generation and editing with flexible duration.

Kling O3

$0.075 - 0.125/s

per second

Advanced multimodal control with per-second billing.

Motion Control

$0.113 - 0.151/s

per second

Motion transfer with per-second billing up to 30 seconds.

Related Kling guides

After you pick a route direction on this family page, use these guides for pricing decisions, O1 evaluation, API access planning, and cross-platform comparison.

Kling AI Model Family FAQ

Everything you need to know about the product and billing.

Kling AI is a family of video generation models developed by Kuaishou Technology. The lineup includes Kling 3.0 for standard generation, Kling 3.0 Turbo for faster short-form generation, Kling O1 for unified creation and editing, Kling O3 (V3 Omni) for advanced multimodal control, and Kling 3.0 Motion Control for motion transfer. Through EvoLink, all five routes are accessible via a single API surface.
Kling AI models cover text-to-video, image-to-video, reference-to-video, video editing, and motion transfer. Key features include per-second billing, support for 720p and 1080p output, flexible video duration from 3 to 30 seconds, and consistent character generation across clips. The exact feature set varies by model version.
EvoLink provides access to five Kling model routes: Kling 3.0, Kling 3.0 Turbo, Kling O1, Kling O3, and Kling 3.0 Motion Control. Each route targets a different production need.
Start with Kling 3.0 Turbo if you need faster 720p/1080p text-to-video or first-frame image-to-video. Start with Kling 3.0 for the standard generation route, O1 when character consistency and editing matter, O3 when reference-to-video or advanced editing matters, and Motion Control for movement transfer.
Yes. EvoLink groups all Kling model routes under one account and API key. You can compare models, switch between them, and decide which version should handle each part of your video workflow.