Gemini 3.6 Flash — Is Google BACK? Big coding test
📥 Collaboration: [email protected]
🚀Main Telegram channel: https://t.me/kostik_neuro
🚀Purchase a Codex subscription with a guarantee: https://t.me/kostik_neuro/719
Gemini 3.6 Flash is Google's latest attempt to re-enter the race for the best neural networks for programming.
In this video, I'll run the model through my real-world coding and vibe coding tests. We'll examine the quality of the results, prompt adherence, design, logic, error rate, and how well Gemini can correct its own code.
Can Gemini 3.6 Flash compete with cutting-edge models from OpenAI, Anthropic, Qwen, and other developers? Or is Google a bit late again?
Model specifications and cost:
https://openrouter.ai/google/gemini-3...
In this video, we'll check:
— Code quality and structure
— Compliance with requirements
— Design and responsiveness
— Animations and interactivity
— Project logic
— Errors and self-corrections
— Final result and evaluation
#Gemini #Gemini36Flash #VibeCoding
Timecodes:
Timecodes:
00:00 - Introduction
00:37 - Test 1
02:51 - Test 1. Scores
03:14 - Test 2
05:06 - Test 2. Scores
05:20 - Test 3
08:35 - Test 3. Scores
08:50 - Test 4
12:35 - Test 4. Scores
12:56 - Cost and Results