Enterprise AI Resource Hub - 302.AI | Pay-as-you-go, Comprehensive AI model API Access, Instant Online App Usage

Kling V3(Image-to-video)

Kling flagship image-to-video model

2026-02-06

Video Generation

Pricing:

$0.084/Second

starting from

Bulk order? Contact your manager for exclusive deals

稳定性

Stable

API Overview

Supports synchronous audio generation (sound=True)

Kling V3.0 Image-to-Video is a dual-mode product line for generating videos from images, launched by Kuaishou. It includes two versions: Standard and Pro. Its core positioning is to transform static images into high-quality dynamic videos featuring cinematic camera movements, natural subject motions, and optional synchronized sound effects via a powerful API.

Dual-version strategy: The Pro version delivers top-tier visual fidelity, smoother physics simulations, and stronger adherence to prompt instructions; while the Standard version significantly reduces usage costs while maintaining high availability.
Core capabilities: Both versions support uploading a starting image (image) and optionally an ending image (end_image) to achieve controllable transitions. They also support video lengths of 5 or 10 seconds and multiple aspect ratios (16:9 / 9:16 / 1:1).
Audio integration: Both versions allow enabling sound to generate ambient audio that matches the visuals; the Pro version additionally supports adding up to 2 custom voice entries (voice_list) for character dialogue.
Applicable scenarios: Photo animation, product demonstration videos, social media dynamic cover images, ad material extensions, and story-driven content creation with sound effects.
Cost comparison: The Pro version costs about twice as much as the Standard version. Users can flexibly choose based on their quality requirements.

───────────────────────────────────────────────────────────────────

Core Capabilities

🖼️→🎥 Single-image-driven animation

With just one reference image and a text description, you can generate coherent dynamic videos, dramatically lowering the barrier to video production. 🔄 Start-and-end frame guidance(optional)

By controlling the end image, you can define the final state of the video, enabling semantic transitions such as "flower blooming → flower wilting" or "stillness → running."

🔊 Synchronized audio and video generation

When sound is enabled, ambient audio is automatically added; in the Pro version, you can also overlay custom voices to create composite audio featuring "character dialogue + background sound."

🎬 Cinematic motion performance (Pro) The Pro version excels in subject consistency, light and shadow changes, and smooth camera movements, making it ideal for delivering finished products.

🚫 Precise content control

Negative prompts allow you to exclude unwanted elements such as "blur" or "distortion," enhancing the stability of generated results. ───────────────────────────────────────────────────────────────────

Effect Demonstrations

API Console

API Reference (3)

API Description API Endpoint Request Method Stability Parameter Description

Kling v3-std

https://api.302.ai/ws/api/v3/kwaivgi/kling-v3.0-std/image-to-video

POST

Stable

View Details

Document Details

Kling 3.0-std delivers high-quality image-to-video generation with smooth motion, cinematic visuals, accurate prompt adherence, and native audio ready for sharing clips.

Supports synchronous audio generation (sound=True)

Pricing

Duration	Sound Off	Sound On
3s	0.504 PTC	0.756 PTC
5s	0.84 PTC	1.26 PTC
10s	1.68 PTC	2.52 PTC
15s	2.52 PTC	3.78 PTC

Request Parameters

Header Parameters

Authorizationstring

Optional

Example Value:Bearer {{YOUR_API_KEY}}

Request Body application/json

promptstring

Required

📝 Generate Positive Prompts

Core Function: Define the video’s core content, style, dynamic effects, and visual characteristics
Filling Tip: Combine features from the starting/ending images; the more specific the description, the closer the generated result will match expectations

Example Value:a butterfly flying from a flower to a garden, soft light, smooth motion

negative_promptstring

Optional

🚫 Generate Negative Prompts

Core Function: Exclude unwanted elements, styles, or visual flaws from the video
Recommendation: Clearly specify content to avoid, such as blurry, pixelated, text, watermark, unnatural motion, etc.

Example Value:blurry, pixelated, text, watermark, overexposed, unnatural motion

imagestring

Required

🖼️ Starting Image URL (Required)

Core Function: Serves as the base starting frame for video generation
📋 Format Requirements: Only .jpg/.jpeg/.png formats supported
⚠️ Specifications:
- File size ≤ 10MB
- Width/Height ≥ 300px
- Aspect ratio range: 1:2.5 ~ 2.5:1

Example Value:https://example.com/start_flower.jpg

end_imagestring

Optional

🎞️ End Image URL (Optional)

Core Function: Serves as the reference frame for the ending shot of video generation, enabling smooth visual transitions
Recommendation: Maintain consistent style with the starting image to enhance the naturalness of video transitions

Example Value:https://example.com/end_garden.png

durationenum<integer>

Optional

⏱️ Video Duration (in seconds)

Options: Only 5 seconds and 10 seconds are supported
Default value if not specified: 5

Enum Value:

510

cfg_scalenumber

Optional

🎛️ Video Generation Flexibility Coefficient

Value range: 0.00 ~ 1.00 (rounded to two decimal places)
⚡ Core rules:
- Higher value → lower model flexibility → stronger alignment with prompts/reference images
- Lower value → greater creative freedom for the model → richer visual effects but potentially deviating from expectations
Default value if left unspecified: 0.5

soundboolean

Optional

🔊 Whether to synchronously generate video audio

true: Generate background audio matching the video style (natural sound effects/light music)
false: Generate video files without audio only

Example Value:true

multi_promptarray[string]

Optional

📜 Multiple Prompt List

Core Function: Define visual changes for different segments of the video; prompts take effect sequentially according to array order
Format Requirement: Each element in the array is an independent prompt text, corresponding to the content of different time segments in the video

Example Value:

[
  "butterfly on flower",
  "butterfly flying through air",
  "butterfly arriving at garden"
]

Kling v3-pro

https://api.302.ai/ws/api/v3/kwaivgi/kling-v3.0-pro/image-to-video

POST

Stable

View Details

Document Details

Kling 3.0 Pro delivers top-tier image-to-video generation with smooth motion, cinematic visuals, precise prompt adherence, and native audio ready for sharing clips.

Supports synchronous audio generation (sound=True)

Pricing

Duration	Sound Off	Sound On
3s	0.672 PTC	1.008 PTC
5s	1.12 PTC	1.68 PTC
10s	2.24 PTC	3.36 PTC
15s	3.36 PTC	5.04 PTC

Request Parameters

Header Parameters

Authorizationstring

Optional

Example Value:Bearer {{YOUR_API_KEY}}

Request Body application/json

promptstring

Required

📝 Generate Positive Prompts

Core Function: Define the video’s core content, style, and dynamic transition effects
Filling Tip: Combine the visual characteristics of the starting/ending images; the more specific the description, the more natural the transition effect

Example Value:smooth transition from a single rose to a bouquet, soft focus, gentle movement

negative_promptstring

Optional

🚫 Generate Negative Prompts

Core Function: Exclude unwanted elements, styles, or visual flaws from the video
Filling Tip: Clearly specify content to avoid, such as blurry, pixelated, text, watermark, distorted aspect ratio, etc.

Example Value:blurry, pixelated, text, watermark, distorted aspect ratio, overexposed colors

imagestring

Required

🖼️ Starting Image URL (Required)

Core Function: Serves as the base starting frame for video generation
📋 Format Requirements: Only .jpg/.jpeg/.png formats supported
⚠️ Strict Specifications:
- File Size: ≤ 10MB
- Dimension Requirement: Width/Height ≥ 300px
- Aspect Ratio Range: 1:2.5 ~ 2.5:1 (content outside this range cannot be processed)

Example Value:https://example.com/start_rose.jpg

end_imagestring

Optional

🎞️ End Image URL (Optional)

Core Function: Serves as the reference frame for the ending shot of video generation, enabling a smooth transition from start to end.
Recommendation: Maintain consistent format/specifications with the starting image to enhance transition smoothness.

Example Value:https://example.com/end_bouquet.png

durationenum<integer>

Optional

⏱️ Video Duration (in seconds)

Options: Only 5 seconds and 10 seconds are supported
Default value if not specified: 5

Enum Value:

510

cfg_scalenumber

Optional

🎛️ Video Generation Flexibility Coefficient

Value range: 0.00 ~ 1.00 (rounded to two decimal places)
⚡ Core rules:
- Higher value → Lower model flexibility → Stronger alignment with prompts/reference images
- Lower value → Greater creative freedom for the model → Richer visual effects but potentially deviating from expectations
Default value if left blank: 0.5

soundboolean

Optional

🔊 Whether to synchronously generate video audio

true: Generate background audio matching the video style (natural sound effects/light music)
false: Generate video files without audio only

Example Value:true

multi_promptarray[string]

Optional

📜 Multiple Prompt List

Core Function: Define visual changes for different time segments of the video; prompts take effect sequentially according to array order
Format Requirement: Each element in the array is an independent prompt text, corresponding to the content of different time segments in the video

Example Value:

[
  "single rose blooming",
  "petals spreading",
  "forming a full bouquet"
]

Fetch

https://api.302.ai/ws/api/v3/predictions/{requestId}/result

GET

Stable

View Details

Document Details

Usage:
Obtain the requestId from the request of a generation task and fill it into the API to retrieve the result.

Price: Free

Request Parameters

Path Parameters

requestIdstring

Required

Header Parameters

Authorizationstring

Required

Example Value:Bearer {{YOUR_API_KEY}}

API Pricing

$￥円 ₽

Model	Description	302.AI Price

Kling v3-std	No synchronous sound is generated	$0.084/Second
Kling v3-std	Generate synchronous sound	$0.126/Second
Kling v3-pro	No synchronous sound is generated	$0.112/Second
Kling v3-pro	Generate synchronous sound	$0.168/Second
Fetch	Fetch Task	Free

Kling V3(Image-to-video)

API Overview

Core Capabilities

Effect Demonstrations

API Console

API Reference (3)

Header Parameters

Request Body application/json

Header Parameters

Request Body application/json

Path Parameters

Header Parameters

API Pricing

Kling v3-std

$0.084/Second

Kling v3-std

$0.126/Second

Kling v3-pro

$0.112/Second

Kling v3-pro

$0.168/Second

Fetch

Free

Recommended

Quick Access

Support

Legal Statement