Kling Video Skill

An open AI agent skill for generating videos using Kling AI via Atlas Cloud API

Generate stunning AI videos directly from your terminal or through AI coding agents (Claude Code, Cursor, Codex, Copilot, Gemini CLI, Windsurf, and more). Powered by Kling AI's state-of-the-art video generation models, accessible through Atlas Cloud's unified API.

Features

Text-to-Video (t2v) — Generate videos from text prompts with cinematic quality
Image-to-Video (i2v) — Animate still images into dynamic video sequences
Reference-to-Video (ref) — Use reference videos to guide style, motion, and composition
Video Editing (edit) — Modify existing videos with text instructions
Avatar Generation — Create talking head videos from a single portrait image
Motion Control — Fine-grained control over camera movement and subject motion
Multi-Shot Generation — Generate multi-scene videos with consistent characters and settings
4K Support — Output videos in resolutions up to 4096x4096
Native Audio Generation — Generate synchronized audio tracks alongside video
Agent Skill — Works with 15+ AI coding agents including Claude Code, Cursor, OpenAI Codex, GitHub Copilot, Gemini CLI, Windsurf, OpenCode, Kiro, and more

Prerequisites

Bun runtime (v1.0+)
An Atlas Cloud API key

Installation

1. Clone the Repository

git clone https://github.com/thoughtincode/kling-video-skill.git
cd kling-video-skill

2. Install Dependencies

bun install

3. Link the CLI Globally

bun link

This makes the kling-video command available system-wide.

4. Set Your API Key

export ATLASCLOUD_API_KEY=your_api_key_here

Or create a .env file in the project root:

ATLASCLOUD_API_KEY=your_api_key_here

Agent Skill Integration

This repository is designed to work as an open agent skill compatible with 15+ AI coding agents. Install it once and use it across platforms.

Supported Agents

Claude Code, Cursor, OpenAI Codex, GitHub Copilot, Gemini CLI, Windsurf, OpenCode, Kiro, and more.

How It Works

Clone this repo and run bun link
Load the skill in your agent (e.g., run /init in Claude Code, or follow your agent's skill loading process)
Use natural language to generate videos:

"Generate a 5-second video of a sunset over the ocean with gentle waves"

"Create a video from this image: ./photo.jpg — make the clouds move slowly"

"Make a 10-second cinematic video of a futuristic city at night using Kling v3.0 Pro"

Your AI coding agent will automatically construct and execute the appropriate kling-video CLI command based on your natural language request.

CLI Usage

Basic Text-to-Video

kling-video "A majestic eagle soaring through mountain valleys at golden hour"

Image-to-Video

kling-video "The flowers gently sway in the breeze" --mode i2v --image ./flower.jpg

Specify Model and Duration

kling-video "Cyberpunk city street with neon reflections in rain puddles" \
  --model kwaivgi/kling-v3.0-pro/text-to-video \
  --duration 10

Full Options Example

kling-video "A time-lapse of storm clouds forming over a desert landscape" \
  --model kwaivgi/kling-v3.0-pro/text-to-video \
  --mode t2v \
  --duration 10 \
  --resolution 1080p \
  --aspect-ratio 16:9 \
  --audio \
  --output ./storm-timelapse.mp4

Reference-to-Video

kling-video "Recreate this scene with autumn colors" \
  --mode ref \
  --image ./reference-scene.mp4

Video Editing

kling-video "Add dramatic lightning flashes to the sky" \
  --mode edit \
  --image ./original-video.mp4

Model Variants

Kling AI offers multiple model variants through Atlas Cloud, each optimized for different use cases:

Model	Model ID	Starting Price per Second	Speed	Quality	Best For
Kling v3.0 Pro	`kwaivgi/kling-v3.0-pro/text-to-video`	from $0.204/s	Medium	Highest	Production-quality cinematic videos
Kling v3.0 Standard	`kwaivgi/kling-v3.0-std/text-to-video`	from $0.102/s	Fast	High	Balanced quality and speed
Kling v2.6 Pro	`kwaivgi/kling-v2.6-pro/text-to-video`	from $0.168/s	Medium	Very High	Proven reliability, complex scenes
Kling v2.5 Turbo Pro	`kwaivgi/kling-v2.5-turbo-pro/text-to-video`	from $0.084/s	Very Fast	Good	Rapid iteration, previews
Kling O3 Pro	`kwaivgi/kling-o3-pro/text-to-video`	from $0.204/s	Medium	Highest	Optimized motion, smooth animation
Kling O1	`kwaivgi/kling-o1/text-to-video`	from $0.102/s	Fast	High	General purpose, cost-effective

Image-to-Video Variants

Model	Model ID	Starting Price per Second
Kling v3.0 Pro i2v	`kwaivgi/kling-v3.0-pro/image-to-video`	from $0.204/s
Kling v3.0 Std i2v	`kwaivgi/kling-v3.0-std/image-to-video`	from $0.102/s
Kling v2.6 Pro i2v	`kwaivgi/kling-v2.6-pro/image-to-video`	from $0.168/s
Kling O3 Pro i2v	`kwaivgi/kling-o3-pro/image-to-video`	from $0.204/s

Default model: kwaivgi/kling-v3.0-pro/text-to-video

Options Reference

Option	Flag	Default	Description
Prompt	(positional)	required	Text description of the video to generate
Model	`--model`	`kwaivgi/kling-v3.0-pro/text-to-video`	Model variant to use (see table above)
Mode	`--mode`	`t2v`	Generation mode: `t2v`, `i2v`, `ref`, `edit`
Duration	`--duration`	`5`	Video duration in seconds (5, 10)
Resolution	`--resolution`	`1080p`	Output resolution: `480p`, `720p`, `1080p`, `4k`
Aspect Ratio	`--aspect-ratio`	`16:9`	Video aspect ratio: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `21:9`
Audio	`--audio`	`false`	Enable native audio generation
Image	`--image`	—	Input image path for i2v/ref/edit modes
Output	`--output`	`./output/{timestamp}.mp4`	Output file path

API Details

This CLI interacts with the Atlas Cloud API to submit video generation jobs and poll for results.

Workflow

Submit Job — POST request to create a prediction
Poll Status — GET request to check prediction status every 5 seconds
Download — Fetch the completed video and save locally

Endpoints

POST   https://api.atlascloud.ai/api/v1/model/prediction
GET    https://api.atlascloud.ai/api/v1/model/prediction/{request_id}

Authentication

All requests require the ATLASCLOUD_API_KEY header:

Authorization: Bearer YOUR_API_KEY

Request Body Example (Text-to-Video)

{
  "model": "kwaivgi/kling-v3.0-pro/text-to-video",
  "input": {
    "prompt": "A cinematic shot of a dragon flying over a medieval castle",
    "duration": 5,
    "resolution": "1080p",
    "aspect_ratio": "16:9",
    "audio_generation": false
  }
}

Request Body Example (Image-to-Video)

{
  "model": "kwaivgi/kling-v3.0-pro/image-to-video",
  "input": {
    "prompt": "The clouds move slowly across the sky",
    "image_url": "https://example.com/image.jpg",
    "duration": 5,
    "resolution": "1080p",
    "aspect_ratio": "16:9"
  }
}

Response Example

{
  "request_id": "abc123-def456",
  "status": "completed",
  "output": {
    "video_url": "https://cdn.atlascloud.ai/videos/abc123.mp4"
  },
  "cost": {
    "amount": 0.204,
    "currency": "USD"
  }
}

Examples

1. Product Demo Video

kling-video "A sleek smartphone rotating 360 degrees on a dark reflective surface, \
studio lighting with subtle color accents, premium product showcase" \
  --duration 5 --resolution 1080p --aspect-ratio 9:16

2. Nature Scene

kling-video "Aerial drone shot of autumn forest with a winding river, \
morning mist rising through golden and red canopy, cinematic color grading" \
  --duration 10 --resolution 1080p --aspect-ratio 21:9

3. Social Media Content

kling-video "A barista doing latte art in slow motion, \
warm cafe ambience, shallow depth of field, Instagram aesthetic" \
  --duration 5 --aspect-ratio 9:16 --audio

4. Animate a Photo

kling-video "The subject smiles and turns their head slightly to the right, \
wind blows through their hair" \
  --mode i2v --image ./portrait.jpg --duration 5

5. Cinematic Landscape

kling-video "Epic wide shot of volcanic eruption at night, \
lava flowing down the mountainside, lightning in the ash cloud, \
National Geographic documentary style" \
  --model kwaivgi/kling-v3.0-pro/text-to-video \
  --duration 10 --resolution 4k --aspect-ratio 21:9

6. Avatar / Talking Head

kling-video "The person delivers a confident TED talk presentation, \
natural gestures and facial expressions" \
  --mode i2v --image ./headshot.jpg --duration 10 --audio

7. Multi-Shot Sequence

# Scene 1: Establishing shot
kling-video "Wide establishing shot of a cyberpunk megacity, \
flying cars and holographic billboards" \
  --duration 5 --output ./scene1.mp4

# Scene 2: Street level
kling-video "Street-level POV walking through neon-lit alleyways, \
steam rising from vents, holographic street vendors" \
  --duration 5 --output ./scene2.mp4

# Scene 3: Close-up
kling-video "Close-up of a cybernetic eye powering on with blue light, \
reflection of the city visible in the iris" \
  --duration 5 --output ./scene3.mp4

8. Motion-Controlled Camera

kling-video "Smooth dolly zoom through a grand library with infinite bookshelves, \
camera slowly pushing forward while zooming out, Vertigo effect, \
warm ambient lighting with dust particles in the air" \
  --duration 10 --resolution 1080p

9. Video Editing

kling-video "Transform the daytime scene to nighttime, \
add moonlight and stars, turn on the building lights" \
  --mode edit --image ./daytime-city.mp4

10. Batch Generation Script

#!/bin/bash
prompts=(
  "Ocean waves crashing on rocky coastline at sunset"
  "Cherry blossoms falling in slow motion in a Japanese garden"
  "Northern lights dancing over a frozen lake in Iceland"
  "Time-lapse of clouds forming and dissolving over mountains"
)

for i in "${!prompts[@]}"; do
  kling-video "${prompts[$i]}" \
    --duration 5 \
    --resolution 1080p \
    --output "./output/batch_${i}.mp4"
done

Pricing

Kling AI on Atlas Cloud

Model Variant	Starting Price per Second	Duration	Resolution
Kling v3.0 Pro	from $0.204/s	5-10s	Up to 4K
Kling v3.0 Standard	from $0.102/s	5-10s	Up to 1080p
Kling v2.6 Pro	from $0.168/s	5-10s	Up to 1080p
Kling v2.5 Turbo Pro	from $0.084/s	5s	Up to 1080p
Kling O3 Pro	from $0.204/s	5-10s	Up to 4K
Kling O1	from $0.102/s	5-10s	Up to 1080p

Prices shown are starting prices. Higher resolution or longer duration may cost more.

Volume Estimates

Volume	Kling v3.0 Pro	Kling v3.0 Std	Kling v2.5 Turbo
100 videos	$20.40	$10.20	$8.40
500 videos	$102.00	$51.00	$42.00
1,000 videos	$204.00	$102.00	$84.00

First-Time Bonus

Atlas Cloud offers a 25% bonus on your first recharge (up to $100 bonus):

Top-Up	Bonus (25%)	Total Credits	Effective Rate (v3.0 Pro)
$10	$2.50	$12.50	from $0.163/s
$50	$12.50	$62.50	from $0.163/s
$100	$25.00	$125.00	from $0.163/s
$400	$100.00	$500.00	from $0.163/s

Take This to Production Today

This workflow is optimized for Atlas Cloud. Move from experiment to enterprise-ready scale.

Production-Ready: 20+ Kling model variants, from $0.204/s
Enterprise Security: SOC I & II Certified | HIPAA Compliant
Zero Maintenance: Serverless architecture — focus on your product, not the servers
25% Bonus: First recharge bonus up to $100

Start Building

Troubleshooting

Common Issues

ATLASCLOUD_API_KEY not set

Error: ATLASCLOUD_API_KEY environment variable is required

Set the environment variable or create a .env file.

Prediction timeout

Error: Prediction timed out after 300 seconds

Video generation can take 1-5 minutes. The default timeout is 300 seconds. For complex prompts or higher resolutions, the generation may take longer.

Invalid model ID

Error: Model not found

Check the Model Variants table for valid model IDs.

Image file not found (i2v mode)

Error: Image file not found: ./photo.jpg

Ensure the image path is correct and the file exists.

Architecture

kling-video-skill/
├── src/
│   └── cli.ts          # Main CLI entry point
├── output/             # Default output directory for generated videos
├── CLAUDE.md           # Claude Code skill definition
├── package.json        # Package configuration
├── .env.example        # Environment variable template
├── .gitignore          # Git ignore rules
├── LICENSE             # MIT License
└── README.md           # This file

Contributing

Fork the repository
Create your feature branch (git checkout -b feature/new-feature)
Commit your changes (git commit -m 'Add new feature')
Push to the branch (git push origin feature/new-feature)
Open a Pull Request

License

This project is licensed under the MIT License — see the LICENSE file for details.

Built for the open agent skills ecosystem — works with Claude Code, Cursor, Codex, Copilot, Gemini CLI, Windsurf, OpenCode, Kiro, and 15+ AI coding agents. Powered by Atlas Cloud.

Name		Name	Last commit message	Last commit date
Latest commit History 6 Commits
scripts		scripts
src		src
.env.example		.env.example
.gitignore		.gitignore
CLAUDE.md		CLAUDE.md
LICENSE		LICENSE
README.md		README.md
SKILL.md		SKILL.md
package.json		package.json

Folders and files

Latest commit

History

Repository files navigation

Kling Video Skill

Table of Contents

Features

Prerequisites

Installation

1. Clone the Repository

2. Install Dependencies

3. Link the CLI Globally

4. Set Your API Key

Agent Skill Integration

Supported Agents

How It Works

CLI Usage

Basic Text-to-Video

Image-to-Video

Specify Model and Duration

Full Options Example

Reference-to-Video

Video Editing

Model Variants

Image-to-Video Variants

Options Reference

API Details

Workflow

Endpoints

Authentication

Request Body Example (Text-to-Video)

Request Body Example (Image-to-Video)

Response Example

Examples

1. Product Demo Video

2. Nature Scene

3. Social Media Content

4. Animate a Photo

5. Cinematic Landscape

6. Avatar / Talking Head

7. Multi-Shot Sequence

8. Motion-Controlled Camera

9. Video Editing

10. Batch Generation Script

Pricing

Kling AI on Atlas Cloud

Volume Estimates

First-Time Bonus

Take This to Production Today

Troubleshooting

Common Issues

Architecture

Contributing

License

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages