The reasoning modes are really weird with 4.7 In my tests, asking for "none" rea...

Eifert · 2026-04-18T11:10:02 1776510602

Medium reasoning has regressed since 4.6. While None and Max have improved since 4.6 in our benchmark. We suspect that this is how Claude tries to cope with the increased user base. Note, Google and OpenAI probably did something similar long ago.

XCSme · 2026-04-17T00:38:04 1776386284

Oh, and also, the "none" and "medium" variants performed the same (??)

nisarg2 · 2026-04-17T03:21:08 1776396068

Insane! Even Haiku doesn't make such mistakes.

XCSme · 2026-04-17T10:36:38 1776422198

I am not sure it's a mistake, this might be their new "adaptive reasoning" + hidden reasoning trace, so we can't verify.

amelius · 2026-04-17T07:48:29 1776412109

Claude is known for its shitty metering.