Claude 3 Haiku vs Gemini 2.5 Pro Preview 06-05
Compare Claude 3 Haiku by Anthropic against Gemini 2.5 Pro Preview 06-05 by Google AI, context windows of 200K vs 1.0M, tested across 44 shared challenges.
Our Verdict
Gemini 2.5 Pro Preview 06-05
Claude 3 HaikuRunner-up
No community votes yet. On paper, Gemini 2.5 Pro Preview 06-05 has the edge — bigger model tier, newer, bigger context window.
Claude 3 Haiku is 8.0x cheaper per token — worth considering if cost matters.
Slight edge
API pricing
Cost per 1M tokens
Claude 3 Haiku
Input
$0.25
5.0× cheaper
Output
$1.25
8.0× cheaper
Gemini 2.5 Pro Preview 06-05
Input
$1.25
Output
$10.00
Claude 3 Haiku is cheaper on both — 5.0× input, 8.0× output
Writing DNA
Style Comparison
Similarity
51%
Gemini 2.5 Pro Preview 06-05 uses 98.0x more headings
Claude 3 Haiku
Gemini 2.5 Pro Preview 06-05
50%Vocabulary51%
22wSentence Length15w
0.71Hedging0.31
0.3Bold4.8
5.9Lists4.1
0.00Emoji0.00
0.00Headings0.98
0.08Transitions0.05
Based on 13 + 20 text responses
Research
What we learned reading every model
FAQ
