DeepSeek FLASH Destroys Gemini FLASH

Discover AI · Advanced ·🧠 Large Language Models ·2mo ago

Key Takeaways

Demonstrates the performance difference between DeepSeek and Gemini FLASH AI models

Original Description

Simple causal logical test. all other AI model performance test result here https://www.youtube.com/watch?v=wlLoLs-KSSg&list=PLgy71-0-2-F0Rla8lu5ZldpYQUfXM_5bT 00:00 DeepSeek V4 FLASH Thinking 00:56 First Result Gemini Flash 01:45 First Result DeepSeek V4 Flash 01:55 Results 1 02:34 Invalid Gemini Flash 02:57 Valid DeepSeek V4 Flash 03:19 Results 2 03:26 Optimization Run 04:15 Optimized result Gemini 04:54 Optimized result DeepSeek 05:12 Results 3 06:34 Gemini 3.1 Flash Lite HIGH 06:50 Optimization run HIGH 08:01 Result Gemini 3.1 Flash High @deepseek-api @googledeepmind #geminiai #gemini3 #deepseek #deepseek_v4 #deepseekai #aitesting
Watch on YouTube ↗ (saves to browser)
Sign in to unlock AI tutor explanation · ⚡30

Related Reads

📰
Open-Weight LLM API Integration: A Developer's Guide to Building with Transparent AI
Learn to integrate open-weight LLM APIs into your applications for transparent AI, without managing GPU clusters or local model weights
Dev.to AI
 📰
Rater State Bias in RLHF Preference Data: An Audit Framework
Learn to identify and audit rater state bias in RLHF preference data to improve model reliability
ArXiv cs.AI
 📰
Some Large Language Models Exhibit Consistent Risk Attitudes
Discover how large language models exhibit consistent risk attitudes and learn to evaluate their decision-making under uncertainty
ArXiv cs.AI
 📰
A Survey on GNN-based Link Prediction: Techniques, Applications, and Challenges
Learn how Graph Neural Networks (GNNs) enable link prediction in various graph structures and applications, and discover the challenges and techniques involved.
ArXiv cs.AI

Chapters (14)

DeepSeek V4 FLASH Thinking
0:56 First Result Gemini Flash
1:45 First Result DeepSeek V4 Flash
1:55 Results 1
2:34 Invalid Gemini Flash
2:57 Valid DeepSeek V4 Flash
3:19 Results 2
3:26 Optimization Run
4:15 Optimized result Gemini
4:54 Optimized result DeepSeek
5:12 Results 3
6:34 Gemini 3.1 Flash Lite HIGH
6:50 Optimization run HIGH
8:01 Result Gemini 3.1 Flash High
Up next
5 Levels of AI Agents - From Simple LLM Calls to Multi-Agent Systems
Dave Ebbelaar (LLM Eng)
Watch →