Gemini 2.0 Flash Thinking vs GPT-4.5
Compare Gemini 2.0 Flash Thinking by Google AI against GPT-4.5 by OpenAI, context windows of 500K vs 128K, tested across 15 shared challenges.
Our Verdict
GPT-4.5
Gemini 2.0 Flash ThinkingRunner-up
No community votes yet. On paper, GPT-4.5 has the edge — bigger model tier, newer.
Gemini 2.0 Flash Thinking is 300x cheaper per token — worth considering if cost matters.
Too close to call
API pricing
Cost per 1M tokens
Gemini 2.0 Flash Thinking
Input
$0.25
300× cheaper
Output
$0.50
300× cheaper
GPT-4.5
Input
$75.00
Output
$150.00
Gemini 2.0 Flash Thinking is cheaper on both — 300× input, 300× output
Writing DNA
Style Comparison
Similarity
47%
GPT-4.5 uses 38.8x more headings
Gemini 2.0 Flash Thinking
GPT-4.5
54%Vocabulary63%
15wSentence Length16w
0.64Hedging0.57
3.8Bold4.7
2.4Lists6.4
0.00Emoji0.00
0.03Headings1.30
0.30Transitions0.54
Based on 13 + 11 text responses
Research
What we learned reading every model
FAQ
