News for hackers. Stuff that matters.

A comment on Claude: Elevated Error Rates for Opus 4.8, Opus 4.7, Opus 4.6, and Sonnet 4.6. Parent comment

Re:Claude: Elevated Error Rates for Opus 4.8, Opus 4.7, Opus... by mixtureoftakes on Sunday June 21, 2026 @08:55PM (#48624271)
yeah and lets not forget codex and glm have subscriptions too, with even more usage per dollar
[ Parent | on HN ]
Re:Claude: Elevated Error Rates for Opus 4.8, Opus 4.7, Opus... by behnamoh on Sunday June 21, 2026 @08:56PM (#48624279)
but they also burn more tokens per task, so in the end, Claude comes out as the more efficient one, despite giving you less tokens.
[ Parent | on HN ]
Re:Claude: Elevated Error Rates for Opus 4.8, Opus 4.7, Opus... by viraptor on Sunday June 21, 2026 @09:14PM (#48624412)
You've got it backwards. Opus is the token/money burning one https://deepswe.datacurve.ai/

Gpt 5.5 uses a third of the opus 4.8 tokens for the same task and scores higher. Glm 5.2 was worse in quality but used half the tokens - 5.3 is not tested yet but will be higher.

[ Parent | on HN ]