Claude Haiku 4.5 vs GPT OSS 20B
Compare Claude Haiku 4.5 by Anthropic against GPT OSS 20B by OpenAI, context windows of 200K vs 131K, tested across 54 shared challenges.
Our Verdict
Claude Haiku 4.5
GPT OSS 20BRunner-up
No community votes yet. On paper, Claude Haiku 4.5 has the edge — newer, bigger context window.
GPT OSS 20B is 50x cheaper per token — worth considering if cost matters.
Too close to call
API pricing
Cost per 1M tokens
Claude Haiku 4.5
Input
$1.00
Output
$5.00
GPT OSS 20B
Input
$0.02
50× cheaper
Output
$0.10
50× cheaper
GPT OSS 20B is cheaper on both — 50× input, 50× output
Writing DNA
Style Comparison
Similarity
48%
Claude Haiku 4.5 uses 3.3x more headings
Claude Haiku 4.5
GPT OSS 20B
61%Vocabulary54%
47wSentence Length17w
0.64Hedging0.26
5.1Bold5.9
6.3Lists3.3
0.20Emoji0.10
2.46Headings0.75
0.17Transitions0.32
Based on 24 + 21 text responses
Research
What we learned reading every model
FAQ
