Qwen: Qwen3 235B A22B Thinking 2507 vs Qwen: Qwen3.6 27B
Compare Qwen: Qwen3 235B A22B Thinking 2507 and Qwen: Qwen3.6 27B, both from Qwen, context windows of 131K vs 256K, tested across 51 shared challenges.
Our Verdict
Qwen: Qwen3 235B A22B Thinking 2507
Qwen: Qwen3.6 27B
No community votes yet. On paper, these are closely matched - try both with your actual task to see which fits your workflow.
Qwen: Qwen3 235B A22B Thinking 2507 is 5.4x cheaper per token — worth considering if cost matters.
Too close to call
API pricing
Cost per 1M tokens
Qwen: Qwen3 235B A22B Thinking 2507
Input
$0.11
3.0× cheaper
Output
$0.60
5.4× cheaper
Qwen: Qwen3.6 27B
Input
$0.33
Output
$3.25
Qwen: Qwen3 235B A22B Thinking 2507 is cheaper on both — 3.0× input, 5.4× output
Writing DNA
Style Comparison
Similarity
53%
Qwen: Qwen3.6 27B uses 2.9x more emoji
Qwen: Qwen3 235B A22B Thinking 2507
Qwen: Qwen3.6 27B
56%Vocabulary62%
14wSentence Length16w
0.37Hedging0.28
5.7Bold4.3
4.0Lists2.1
0.54Emoji1.54
0.70Headings0.68
0.14Transitions0.05
Based on 22 + 25 text responses
Research
What we learned reading every model
FAQ
