# 🎬 Veo 3.1 MCP - Quick Reference
## 🚀 One-Minute Setup
```bash
cd veo-mcp
npm install
npm run build
echo "GEMINI_API_KEY=your_key_here" > .env
```
Add to `~/.cursor/mcp.json`:
```json
{
"mcpServers": {
"veo": {
"command": "node",
"args": ["C:\\Users\\woute\\Githubs\\MCP\\veo-mcp\\dist\\index.js"],
"env": { "GEMINI_API_KEY": "your_key" }
}
}
}
```
Restart Cursor. Done! ✅
---
## 💬 Common Commands
### Simple Video
```
Generate an 8-second cinematic video of a sunset over mountains
```
### With Reference
```
Create a product showcase video using this style: C:\style.jpg
```
### Frame Interpolation
```
Generate smooth transition between first.jpg and last.jpg, 8 seconds
```
### Check Status
```
Check status of operation: operations/xyz123
```
---
## 📋 Tool Quick Reference
### start_video_generation
**Simple:**
```json
{
"prompt": "A peaceful zen garden at sunrise"
}
```
**Advanced:**
```json
{
"prompt": "Futuristic cityscape at night",
"model": "veo-3.1-fast-generate-001",
"durationSeconds": 8,
"aspectRatio": "16:9",
"resolution": "1080p",
"referenceImages": [{"source": "url", "url": "..."}],
"generateAudio": false
}
```
### get_video_job
```json
{
"operationName": "operations/xyz"
}
```
Poll every 10-30 seconds until `done: true`.
### upload_image
```json
{
"source": "file_path",
"filePath": "C:\\ref.jpg"
}
```
Returns `fileUri` valid 48 hours.
---
## 💰 Quick Pricing
| Duration | Fast (no audio) | Fast (audio) | Quality (no audio) | Quality (audio) |
|----------|-----------------|--------------|-------------------|-----------------|
| 4s | $0.40 | $0.60 | $0.80 | $1.60 |
| 6s | $0.60 | $0.90 | $1.20 | $2.40 |
| 8s | $0.80 | $1.20 | $1.60 | $3.20 |
---
## ⚡ Quick Limits
- Duration: 4, 6, or 8 seconds only
- References: Max 3 images
- Samples: Max 4 per request
- Rate: ~50 requests/min
- File expiry: 48 hours
---
## 🎯 Token Efficiency
### ✅ DO THIS
```json
{
"referenceImages": [{
"source": "url",
"url": "https://example.com/ref.jpg"
}]
}
```
**Cost:** ~20 tokens
### ⚠️ AVOID THIS
```json
{
"referenceImages": [{
"source": "inline",
"base64": "iVBORw0KGgo..." // 500KB
}]
}
```
**Cost:** ~50,000 tokens!
---
## ⏱️ Generation Times
- **4s video:** 30-60 seconds
- **8s video:** 60-120 seconds
- **With audio:** +30-60 seconds
- **Frame interpolation:** +20-40 seconds
---
## 🔧 Troubleshooting
### "403 API not enabled"
```
1. Enable Generative Language API in Cloud Console
2. Enable billing
3. Wait 10 minutes
```
### "Rate limit"
```
Use batch tool with concurrency: 3
Add delays between requests
```
### Videos taking long
```
- Use 720p for testing
- Switch to 1080p for final
- Use fast model for iterations
```
---
## 📊 Async Flow
```
start_video_generation → operationName
↓ (wait 60-120s)
get_video_job → {done: false}
↓ (wait more)
get_video_job → {done: true, videos: [...]}
↓
Download videos
```
---
## 🎬 Pro Tips
1. **Estimate first**: Use `estimate_veo_cost` before batches
2. **Test at 720p**: Switch to 1080p for finals
3. **Pre-upload refs**: Upload once, reuse 48h
4. **Use seeds**: For reproducible variations
5. **Batch wisely**: Max 3-5 concurrent to avoid rate limits
---
## 🔗 Resources
- API Key: https://aistudio.google.com/app/apikey
- Docs: https://ai.google.dev/gemini-api/docs/video
- Pricing: https://cloud.google.com/vertex-ai/generative-ai/pricing
---
**Quick Start → Generate Videos → Build Amazing Content! 🎬**
