Gemini 2.5 Pro Experimental vs Qwen3 30B A3B Thinking 2507
Compare Gemini 2.5 Pro Experimental by Google AI against Qwen3 30B A3B Thinking 2507 by Qwen, context windows of 1.0M vs 262K, tested across 43 shared challenges.
Our Verdict
Gemini 2.5 Pro Experimental
Qwen3 30B A3B Thinking 2507Runner-up
No community votes yet. On paper, Gemini 2.5 Pro Experimental has the edge — bigger model tier, bigger context window, major provider backing.
Qwen3 30B A3B Thinking 2507 is 7.0x cheaper per token — worth considering if cost matters.
Too close to call
API pricing
Cost per 1M tokens
Gemini 2.5 Pro Experimental
Input
$1.00
Output
$2.00
Qwen3 30B A3B Thinking 2507
Input
$0.07
14× cheaper
Output
$0.28
7.0× cheaper
Qwen3 30B A3B Thinking 2507 is cheaper on both — 14× input, 7.0× output
Writing DNA
Style Comparison
Similarity
55%
Qwen3 30B A3B Thinking 2507 uses 67.9x more emoji
Gemini 2.5 Pro Experimental
Qwen3 30B A3B Thinking 2507
54%Vocabulary54%
15wSentence Length14w
0.35Hedging0.12
5.6Bold7.5
3.9Lists4.2
0.00Emoji0.68
0.39Headings0.93
0.17Transitions0.12
Based on 18 + 16 text responses
Research
What we learned reading every model
FAQ
