Google Vertex AI - Studio

📋 Service Information

Provider: Google Vertex AI
Service Type: Web Interface (Vertex AI Studio)
Access URL: https://console.cloud.google.com/vertex-ai/generative
Free Tier: Trial Credits ($300, 90 days)

---

🎯 Service Overview

Vertex AI Studio is Google Cloud’s visual AI testing platform, allowing you to test the powerful capabilities of Gemini models without programming.

Key Advantages:

• 🎨 Visual Interface - No programming required
• 🌟 Gemini 1.5 Pro - 2M ultra-long context
• 🖼️ Multimodal - Text, images, audio, video
• 🎛️ Parameter Adjustment - Customize generation parameters
• 💾 Prompt Management - Save and reuse
• 🏢 Enterprise Features - Complete toolchain

---

🚀 How to Use

Prerequisites

• ✅ Created Google Cloud account
• ✅ Activated $300 trial credits
• ✅ Enabled Vertex AI API

For detailed registration steps, see: Vertex AI Registration Guide

Usage Steps

Step 1: Access Vertex AI Studio

1. Login to https://console.cloud.google.com
2. Search for “Vertex AI”
3. Select “Vertex AI Studio”
4. Or visit directly: https://console.cloud.google.com/vertex-ai/generative

Step 2: Select Model

Choose model in Studio interface:

Model	Context	Use Cases
Gemini 1.5 Pro	2M tokens	Complex tasks, long documents
Gemini 1.5 Flash	1M tokens	Quick response, high frequency

Step 3: Start Testing

1. Enter Prompt
2. Upload files (optional)
3. Adjust parameters (optional)
4. Click “Generate”
5. View results

---

🎨 Feature Highlights

1. Multimodal Testing

Supported Inputs:

• 📝 Text
• 🖼️ Images (JPG, PNG, WEBP)
• 🎵 Audio (MP3, WAV)
• 🎬 Video (MP4, MOV)

Application Scenarios:

• Image recognition and description
• Video content analysis
• Audio transcription
• Multimodal Q&A

2. Ultra-Long Context

Gemini 1.5 Pro:

• 2M tokens context
• Equivalent to 1,400 pages of books
• Analyze complete movie scripts
• Process large volumes of documents

3. Prompt Design

Features:

• Prompt editor
• Parameter adjustment
• Example library
• Best practices

4. Parameter Adjustment

Parameter	Range	Description
Temperature	0-2	Controls creativity
Top P	0-1	Nucleus sampling
Top K	1-40	Number of candidate words
Max Output Tokens	1-8192	Output length

5. Batch Testing

Features:

• Batch prompt testing
• Result comparison
• Performance evaluation

---

🔢 Usage Costs

Based on Trial Credits

Using Vertex AI Studio consumes $300 trial credits:

Gemini 1.5 Pro:

• Input: $1.25/M tokens (≤128K)
• Output: $5/M tokens
• $300 ≈ 240M input tokens

Gemini 1.5 Flash:

• Input: $0.075/M tokens (≤128K)
• Output: $0.30/M tokens
• $300 ≈ 4B input tokens

---

💡 Usage Tips

✅ Best Practices

1. Prioritize Flash

   • Lower cost (16x cheaper)
   • Faster speed
   • Suitable for most scenarios

2. Fully Utilize Long Context

   • Upload complete documents
   • Analyze long conversation history
   • Process large amounts of data

3. Manage Trial Credits

   • Monitor usage
   • Set budget alerts
   • Avoid unexpected consumption

4. Prompt Optimization

   • Use prompt library
   • A/B testing
   • Save best versions

⚠️ Precautions

1. Cost Control

   • Be aware of $300 limit
   • 90-day validity period
   • Monitor consumption rate

2. Network Requirements

   • Needs stable network
   • Mainland China requires VPN
   • Large file uploads take time

3. Data Privacy

   • Be cautious uploading sensitive data
   • Understand data usage policies
   • Use GCP security features

---

🔧 Common Issues

1. How to Check Remaining Credits?

Method:

1. Top of GCP Console
2. Click “Billing”
3. View trial credit balance

2. What Happens After 90 Days?

Explanation:

• No automatic charges after trial ends
• Need to manually upgrade to paid
• Can choose not to upgrade

3. How to Choose Between Pro and Flash?

Suggestions:

• Use Flash for testing (cheaper)
• Use Pro for complex tasks
• Use Pro for long documents (2M context)

---

📚 Related Resources

Documentation and Tutorials

• Vertex AI Complete Guide
• API Service Guide
• Official Documentation

---

Service Provider: Google Vertex AI