Gemini 3 Agentic Coding Test: Is It the NEW KING?

Codewello · Beginner ·🧠 Large Language Models ·4mo ago
Is Google's new model actually the new king of coding? I wanted to find out, so I ran a complete Gemini 3 coding test to see how it handles real development work. In this video, I break down the new features like the Google Deep Think mode and look at the official AI programming benchmark scores to understand the hype. Then, I jump into VS Code (using Copilot) to demo a real Agentic coding workflow. 00:00 - Intro: The Gemini 3 Era 00:15 - Understanding Google Deep Think Mode 00:50 - Gemini 3 New Pricing 01:08 - Temperature Settings 01:39 - Where to Use Gemini 3 02:20 - Gemini 3 Coding App E…
Watch on YouTube ↗ (saves to browser)

Chapters (12)

Intro: The Gemini 3 Era
0:15 Understanding Google Deep Think Mode
0:50 Gemini 3 New Pricing
1:08 Temperature Settings
1:39 Where to Use Gemini 3
2:20 Gemini 3 Coding App Examples
4:16 Benchmarks: Gemini 3 vs GPT
6:50 Gemini 3 Coding Test: VS Code + Copilot
8:00 Building the App: Next.js, Express & Prisma
10:15 Testing Folder Structure & Logic
13:25 The New Google Antigravity IDE
14:30 Learning Design Patterns with Me
5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
Next Up
5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
Dave Ebbelaar (LLM Eng)