deepseek-v4.1POST
Blind test on 15 business emails: DeepSeek V4.1 Flash breaks rules twice, Opus 5.5 none
A blind test of 15 anonymised business emails found DeepSeek V4.1 Flash, Opus 5.5 and the user's own replies all solid, but DeepSeek used a banned greeting twice and invented a promise while Opus broke no rules.
Kevin Theo
@kevintheooArchived 2 days agoOriginally posted 2026.10.26
Kevin Theo @kevintheooOriginally posted 2026.10.26
@EnvolDev different test, same direction. 15 real anonymised business emails, blind: DeepSeek V4.1 Flash vs Opus 5.5 vs my own replies. all pretty good. the only gap was rules: deepseek used a banned greeting twice and invented one promise, opus broke none
x.com/kevintheoo/status/2114440455433973977Archived 2026.10.26 12:03