Kling AI Video Nodes

This library provides Griptape Nodes for interacting with the Kling AI video generation services. You can use these nodes to generate videos from text prompts, images, extend existing videos, or create lip-sync videos.

IMPORTANT: To use these nodes, you will need API keys from Kling AI. Please visit the Kling AI website for more information on how to obtain your keys.

To configure your keys within the Griptape Nodes IDE:

1. Open the Settings menu.
2. Navigate to the API Keys & Secrets panel.
3. Add a new secret configuration for the service named Kling.
4. Enter your KLING_ACCESS_KEY and KLING_SECRET_KEY in the respective fields.

Below is a description of each node and its parameters.

Kling AI Text to Video (KlingAI_TextToVideo)

Generates a video from a text prompt.

Parameter	Type	Description	Default Value
prompt	str	Text prompt for video generation (max 2500 chars).
model_name	str	Model Name.	kling-v1
negative_prompt	str	Negative text prompt (max 2500 chars).	""
cfg_scale	float	Flexibility in video generation (0-1). Higher value = lower flexibility, stronger prompt relevance.	0.5
mode	str	Video generation mode (std: Standard, pro: Professional).	std
aspect_ratio	str	Aspect ratio of the generated video frame (width:height). Choices: 16:9, 9:16, 1:1.	16:9
duration	str	Video Length, unit: s (seconds). Choices: 5, 10.	5
num_videos	int	Number of videos to generate in parallel (1-5).	1
callback_url	str	Callback notification address for task status changes.	""
external_task_id	str	Customized Task ID (must be unique within user account).	""
camera_control_type	str	Predefined camera movement type. Select (Auto) for model default. If simple, ensure only one config value is non-zero. Choices: (Auto), simple, down_back, forward_up, right_turn_forward, left_turn_forward.	(Auto)
camera_config_horizontal	float	Horizontal movement (-10 to 10). Use if camera_control_type is simple.	0.0
camera_config_vertical	float	Vertical movement (-10 to 10). Use if camera_control_type is simple.	0.0
camera_config_pan	float	Pan (rotation around x-axis, -10 to 10). Use if camera_control_type is simple.	0.0
camera_config_tilt	float	Tilt (rotation around y-axis, -10 to 10). Use if camera_control_type is simple.	0.0
camera_config_roll	float	Roll (rotation around z-axis, -10 to 10). Use if camera_control_type is simple.	0.0
camera_config_zoom	float	Zoom (-10 to 10). Use if camera_control_type is simple.	0.0
video_url	VideoUrlArtifact	Output: Video URL (index 0).	None
video_url_1	VideoUrlArtifact	Output: Video URL (index 1).	None
video_url_2	VideoUrlArtifact	Output: Video URL (index 2).	None
video_url_3	VideoUrlArtifact	Output: Video URL (index 3).	None
video_url_4	VideoUrlArtifact	Output: Video URL (index 4).	None
video_urls	list[VideoUrlArtifact]	Output: List of generated videos (completion order).	[]

Note: Callback and Camera Controls parameters are grouped and may be collapsed by default in the UI.

Multiple video generation:
Set num_videos to generate multiple videos in parallel. The first completed video is exposed in video_url, while all completed videos are returned in video_urls in completion order. When num_videos is 1, video_urls contains a single item matching video_url.

Kling AI Image to Video (KlingAI_ImageToVideo)

Generates a video from a reference image and optional text prompts.

Parameter	Type	Description	Default Value
image	ImageArtifact / str	Reference Image (required). Input ImageArtifact, ImageUrlArtifact, direct URL string, or Base64 string.
image_tail	ImageArtifact / str	Reference Image - End frame control. Input ImageArtifact, ImageUrlArtifact, direct URL string, or Base64 string.	None
prompt	str	Positive text prompt (max 2500 chars).	""
negative_prompt	str	Negative text prompt (max 2500 chars).	""
model_name	str	Model Name for generation. Choices: kling-v1, kling-v1-5, kling-v1-6.	kling-v1
cfg_scale	float	Flexibility (0-1). Higher value = lower flexibility, stronger prompt relevance.	0.5
mode	str	Video generation mode (std: Standard, pro: Professional).	std
duration	str	Video Length in seconds. Choices: 5, 10.	5
num_videos	int	Number of videos to generate in parallel (1-5).	1
static_mask	ImageArtifact / str	Static Brush Application Area. Input ImageArtifact, ImageUrlArtifact, direct URL, or Base64 string. Mutually exclusive with Camera Controls.	None
dynamic_masks	str	JSON string for Dynamic Brush Configuration List. Masks within JSON must be URL/Base64. Mutually exclusive with Camera Controls.	None
camera_control_type	str	Predefined camera movement. (Auto) for model default. simple requires one config value. Mutually exclusive with Masks. Choices: (Auto), simple, down_back, forward_up, right_turn_forward, left_turn_forward.	(Auto)
camera_config_horizontal	float	Camera horizontal movement (-10 to 10). Use if type is simple.	0.0
camera_config_vertical	float	Camera vertical movement (-10 to 10). Use if type is simple.	0.0
camera_config_pan	float	Camera pan movement (-10 to 10). Use if type is simple.	0.0
camera_config_tilt	float	Camera tilt movement (-10 to 10). Use if type is simple.	0.0
camera_config_roll	float	Camera roll movement (-10 to 10). Use if type is simple.	0.0
camera_config_zoom	float	Camera zoom movement (-10 to 10). Use if type is simple.	0.0
callback_url	str	Callback notification address for task status changes.	""
external_task_id	str	Customized Task ID (must be unique within user account).	""
video_url	VideoUrlArtifact	Output: Output URL of the generated video (index 0).	None
video_url_1	VideoUrlArtifact	Output: Output URL of the generated video (index 1).	None
video_url_2	VideoUrlArtifact	Output: Output URL of the generated video (index 2).	None
video_url_3	VideoUrlArtifact	Output: Output URL of the generated video (index 3).	None
video_url_4	VideoUrlArtifact	Output: Output URL of the generated video (index 4).	None
video_urls	list[VideoUrlArtifact]	Output: List of generated videos (completion order).	[]

Note: Image Inputs, Prompts, Generation Settings, Masks, Camera Controls, and Callback parameters are grouped and may be collapsed by default in the UI.

Multiple video generation:
Set num_videos to generate multiple videos in parallel. The first completed video is exposed in video_url, while all completed videos are returned in video_urls in completion order. When num_videos is 1, video_urls contains a single item matching video_url.

Kling V3 Multi-Shot (KlingV3MultiShot)

Generates multi-shot video sequences using Kling v3's multi_prompt API. Each shot in the sequence has its own prompt and duration, allowing fine-grained control over the narrative structure of the generated video. An optional start frame and end frame image can guide the visual style.

The shot list is edited through a custom interactive widget (MultiShotEditor) embedded directly in the node. Each shot entry has a name, duration (in seconds), and a text description that serves as the prompt for that segment. The total duration across all shots must be between 3 and 15 seconds, and at least one shot must have a description.

Parameter	Type	Description	Default Value
start_frame	ImageArtifact / ImageUrlArtifact	Starting frame image for the video sequence (required).
end_frame	ImageArtifact / ImageUrlArtifact	Ending frame image (optional).	None
shots	list	List of shots, each with name, duration, and description. Edited via the MultiShotEditor widget.	1 shot, 3s
cfg_scale	float	Flexibility (0-1). Higher value = lower flexibility, stronger prompt relevance.	0.5
mode	str	Video generation mode (std: Standard, pro: Professional).	std
aspect_ratio	str	Aspect ratio of the generated video. Choices: 16:9, 9:16, 1:1.	16:9
negative_prompt	str	Negative text prompt — elements to avoid (max 2500 chars).	""
sound	str	Generate native audio with the video. Choices: on, off.	off
polling_delay	int	Delay in seconds between polling the Kling API for job completion (hidden, 5-30s).	10
video_url	VideoUrlArtifact	Output: URL of the generated multi-shot video.	None

Note: Image Inputs and Generation Settings parameters are grouped and may be collapsed by default in the UI.

Shot list constraints:

• Total duration across all shots must be between 3 and 15 seconds.
• At least one shot must have a non-empty description.
• The node automatically selects the image2video or text2video API endpoint depending on whether a start frame image is provided.

Kling AI Video Extension (KlingAI_VideoExtension)

Extends an existing Kling AI video.

Parameter	Type	Description	Default Value
video_id	str	Required. The ID of the Kling video to extend.
prompt	str	Optional. Text prompt for the extension (max 2500 chars).	""
negative_prompt	str	Optional. Negative text prompt (max 2500 chars).	""
cfg_scale	float	Optional. Flexibility (0-1). Higher value = lower flexibility, stronger prompt relevance.	0.5
callback_url	str	Optional. Callback notification address for task status changes.	""
external_task_id	str	Optional. Customized Task ID for user tracking.	""
extended_video_url	VideoUrlArtifact	Output: URL of the extended video segment.	None
extended_video_id	str	Output: ID of the newly generated extended video segment.	None
extension_task_id	str	Output: Task ID for the video extension job.	None

Note: Core Input, Prompts, Generation Settings, and Callback parameters are grouped and may be collapsed by default in the UI.

Kling AI Lip Sync (KlingAI_LipSync)

Creates lip-sync videos by synchronizing speech to video. Supports both Kling AI generated videos (via video ID) and uploaded videos (via video URL).

Parameter	Type	Description	Default Value
video_input_type	str	Video input type. Choices: video_id (for Kling AI generated videos), video_url (for uploads).	video_id
video_id	str	Video ID from previous Kling AI video generation (required when using video_id input type).
video_url	VideoUrlArtifact / str	Video file or URL for lip sync (required when using video_url input type).
voice_type	str	Voice input type. Choices: text (text-to-speech), audio (audio file).	text
voice_text	str	Text to convert to speech (used when voice_type is text).	""
voice_audio_url	str	URL to audio file (used when voice_type is audio).	""
voice_speaker	str	TTS voice speaker (used when voice_type is text). Choices: ai_shatang, ai_kaiya, ai_chenjiahao_712, ai_huangzhong_712, ai_huangyaoshi_712, ai_laoguowang_712, uk_boy1, uk_man2, uk_oldman3, oversea_male1, commercial_lady_en_f-v1, reader_en_m-v1.	ai_shatang
voice_speed	float	Speech speed multiplier (0.5-2.0). 1.0 = normal speed.	1.0
voice_volume	float	Speech volume multiplier (0.1-2.0). 1.0 = normal volume.	1.0
callback_url	str	Callback notification address for task status changes.	""
lip_sync_video_url	VideoUrlArtifact	Output: URL of the lip-synced video.	None
task_id	str	Output: The Task ID of the lip-sync video from Kling AI.	None

Note: Video Input, Voice Settings, Text-to-Speech Settings, and Callback parameters are grouped and may be collapsed by default in the UI.

Kling AI Motion Control (KlingAI_MotionControl)

Transfers motion from a reference video to a reference image to create a new video.

Parameter	Type	Description	Default Value
prompt	str	Optional text prompt for additional motion guidance (max 2500 chars).	""
reference_image	ImageArtifact / ImageUrlArtifact	Reference image with character (required). Supports .jpg/.jpeg/.png, max 10MB.
reference_video	VideoUrlArtifact	Reference video with actions to transfer (required). Supports .mp4/.mov, max 100MB.
keep_original_sound	bool	Keep original video sound.	True
character_orientation	str	Orientation: image (max 10s video) or video (max 30s video).	video
mode	str	Video generation mode (std: Standard, pro: Professional).	pro
generation_id	str	Output: Kling task ID.	""
provider_response	dict	Output: Verbatim response from API (latest polling response).	None
video_url	VideoUrlArtifact	Output: Saved video URL.	None
kling_video_id	str	Output: The generated video ID from Kling AI.	""
was_successful	bool	Output: Whether the generation succeeded.
result_details	str	Output: Details about the generation result or error.

Note: Image Input, Video Input, Generation Settings, and Status parameters are grouped and may be collapsed by default in the UI.

Add your library to your installed Engine!

If you haven't already installed your Griptape Nodes engine, follow the installation steps HERE. After you've completed those and you have your engine up and running:

1. Copy the path to your griptape_nodes_library.json file within this kling directory. Right click on the file, and Copy Path (Not Copy Relative Path).
2. Start up the engine!
3. Navigate to settings.
4. Open your settings and go to the App Events tab. Add an item in Libraries to Register.
5. Paste your copied griptape_nodes_library.json path from earlier into the new item.
6. Exit out of Settings. It will save automatically!
7. Open up the Libraries dropdown on the left sidebar.
8. Your newly registered library should appear! Drag and drop nodes to use them!