TensorsLab Video Generation

Overview

This skill enables AI-powered video generation through TensorsLab's API, supporting both text-to-video and image-to-video workflows. Video generation is a time-intensive process - tasks typically take several minutes to complete.

Authentication Check

Before any video generation, verify the API key is configured:

CODEBLOCK0

If not set, display this friendly message:

CODEBLOCK1

Models

Default: INLINECODE0

Workflow

1. Text-to-Video Generation

User request: "做一段 10 秒钟横屏的宇宙飞船穿梭星际的视频"

Agent processing:

1. 1. Extract parameters: duration=10, INLINECODE2
2. Enhance prompt with cinematic details, camera movements, scene descriptions
3. Call API with enriched prompt
4. Monitor progress with heartbeat updates (every 60 seconds)
5. Download to INLINECODE3

Example enhanced prompt:
CODEBLOCK2

2. Image-to-Video Generation

User request: "让这张人物合影 family.jpg 动起来" or "让风景照动起来"

Agent processing:

1. 1. Extract image file paths (1-2 images supported)
2. Enhance prompt with motion instructions
3. Monitor progress with heartbeat updates
4. Download results

Parameters for image-to-video:

• - sourceImage: Array of image files (1-2 images max)
• INLINECODE5: Comma-separated URLs of source images
• INLINECODE6: Description of desired motion/animation

3. Resolution and Aspect Ratio

Aspect ratios:

• - 9:16 - Vertical (TikTok, Reels, Shorts) - default
• INLINECODE8 - Horizontal (YouTube, standard video)
• Other ratios available depending on model

Resolutions:

• - 480p - SD quality, faster generation
• INLINECODE10 - HD quality - default
• INLINECODE11 - Full HD
• INLINECODE12 - 2K quality (seedancev2 only)

4. Duration Options

• - seedancev2: 5-15 seconds
• Other models: 5-10 seconds

Longer videos take proportionally more time to generate.

5. Special Features (seedancev2 only)

Progress Tracking

Video generation takes several minutes. Keep users informed:

CODEBLOCK3

Heartbeat interval: Print encouraging message every 60 seconds.

Using the Script

依赖：脚本需要 requests 库，首次使用前执行：
CODEBLOCK4

Execute the Python script directly:

CODEBLOCK5

Task Status Flow

Error Handling

Translate API errors to user-friendly messages:

Output

All videos are saved to output directory with naming pattern:

• - Default: ./tensorslab_output/ (current working directory)
• Custom: Use --output-dir or -o to specify a different path
• Naming: {task_id}_{index}.mp4 - e.g., INLINECODE20

After completion, inform user:
CODEBLOCK6

Tips for Better Results

Text-to-Video

• - Include cinematic terms: "wide shot", "close-up", "pan", "dolly"
• Describe motion: "flying rapidly", "slowly drifting", "zooming in"
• Specify style: "cinematic", "documentary style", "dreamy"

Image-to-Video

• - Describe the desired motion: "gentle sway", "subtle movement"
• For landscapes: "clouds moving", "water flowing", "leaves rustling"

Resources

• - scripts/tensorslabvideo.py: Main API client with full CLI support
• references/apireference.md: Detailed API documentation