veo skill

safe

This skill helps you generate video clips using Google's Veo API, producing MP4 output with customizable duration and aspect ratio.

npx playbooks add skill openclaw/skills --skill veo

Review the files below or copy the command above to add this skill to your agents.

Files (3)

SKILL.md

1.1 KB

---
name: veo
description: Generate video using Google Veo (Veo 3.1 / Veo 3.0).
metadata:
  {
    "openclaw":
      {
        "emoji": "🎬",
        "requires":
          {
            "env": { "GEMINI_API_KEY": "" },
            "bins": ["uv"],
          },
      },
  }
---

# Veo (Google Video Generation)

Generate video clips using Google's Veo API.

Generate video
```bash
uv run {baseDir}/scripts/generate_video.py --prompt "your video description" --filename "output.mp4"
```

Options
- `--duration` / `-d`: Video duration in seconds (default: 8, max varies by model)
- `--aspect-ratio` / `-a`: Aspect ratio (16:9, 9:16, 1:1)
- `--model`: Veo model to use (veo-2.0-generate-001, veo-3.0-generate-001, veo-3.1-generate-preview, etc.)

API key
- `GEMINI_API_KEY` env var (preferred)
- Or set `skills."veo".env.GEMINI_API_KEY` in `~/.clawdbot/clawdbot.json`

Notes
- Veo 3.1 supports higher quality and longer durations
- Output is MP4 format
- Use `--model veo-3.1-generate-preview` for best results
- Veo 3.0-fast-generate-001 is faster but lower quality
- The script prints a `MEDIA:` line for Clawdbot to auto-attach on supported chat providers.

Overview

This skill generates short video clips using Google's Veo video-generation models (Veo 3.1 / 3.0 / 2.0). It provides a command-line script to create MP4 outputs from text prompts and supports adjustable duration, aspect ratio, and model selection. The skill is optimized for quick experimentation and archival of generated outputs.

How this skill works

The tool calls the Veo generation API with your text prompt and model parameters, then streams the returned media into an MP4 file. It supports environment-based API key configuration and prints a MEDIA: line for automated attachment in compatible chat systems. Different model choices trade off speed, quality, and maximum duration.

When to use it

Create short narrated visual clips or concept videos from text prompts
Prototype video ideas before full production
Generate social media assets in different aspect ratios
Batch-generate archival video samples for testing or demos
Quickly compare output quality across Veo model versions

Best practices

Provide clear, concise prompts that describe visual elements and motion
Choose veo-3.1-generate-preview for highest quality and longer durations when available
Use veo-3.0-fast-generate-001 for rapid iteration when short turnaround is needed
Limit duration to the model's supported maximum to avoid errors
Store your GEMINI_API_KEY securely in environment variables rather than embedding it in files

Example use cases

Generate an 8-second product teaser in 16:9 using a descriptive prompt
Create a vertical 9:16 clip for mobile social platforms with veo-3.1-generate-preview
Produce multiple quality variants by switching between veo-3.0-fast-generate-001 and veo-3.1 for comparison
Archive generated MP4 outputs and metadata for later review or demonstration

FAQ

How do I provide the API key?

Set GEMINI_API_KEY in your environment or configure skills."veo".env.GEMINI_API_KEY in your local clawdbot config file.

What model should I pick for best results?

Use veo-3.1-generate-preview for best quality and longer durations; use veo-3.0-fast-generate-001 if you need faster, lower-cost iterations.