Gemini 2.5 Flash Preview 05-20 (thinking) vs Google: Gemini 2.5 Flash Preview 09-2025
Compare Gemini 2.5 Flash Preview 05-20 (thinking) and Google: Gemini 2.5 Flash Preview 09-2025, both from Google AI, context windows of 1.0M vs 1.0M, tested across 12 shared challenges.
Our Verdict
Google: Gemini 2.5 Flash Preview 09-2025
Gemini 2.5 Flash Preview 05-20 (thinking)Runner-up
No community votes yet. On paper, Google: Gemini 2.5 Flash Preview 09-2025 has the edge — bigger model tier, newer.
Too close to call
API pricing
Cost per 1M tokens
Gemini 2.5 Flash Preview 05-20 (thinking)
Input
$0.15
2.0× cheaper
Output
$3.50
Google: Gemini 2.5 Flash Preview 09-2025
Input
$0.30
Output
$2.50
1.4× cheaper
Gemini 2.5 Flash Preview 05-20 (thinking) wins input (2.0× cheaper)·Google: Gemini 2.5 Flash Preview 09-2025 wins output (1.4× cheaper)
Writing DNA
Style Comparison
Similarity
32%
Google: Gemini 2.5 Flash Preview 09-2025 uses 29.7x more hedging
Gemini 2.5 Flash Preview 05-20 (thinking)
Google: Gemini 2.5 Flash Preview 09-2025
70%Vocabulary53%
11wSentence Length19w
0.00Hedging0.30
4.2Bold5.8
7.3Lists3.1
0.00Emoji0.14
0.00Headings0.93
0.00Transitions0.22
Based on 3 + 22 text responses
Research
What we learned reading every model
FAQ
