News for hackers. Stuff that matters.

A comment on Claude 3.7 Sonnet and Claude Code. Parent comment

Re:Claude 3.7 Sonnet and Claude Code by doctoboggan on Monday February 24, 2025 @10:53PM (#43167970)
Have you tried Claude 3.7 + Deepseek as the architect? Seeing as "DeepSeek R1 + claude-3-5-sonnet-20241022" is the second place option, "DeepSeek R1 + claude-3-7" would hopefully be the highest ranking choice so far?
[ Parent | on HN ]
Re:Claude 3.7 Sonnet and Claude Code by SparkyMcUnicorn on Tuesday February 25, 2025 @12:14AM (#43168426)
[delayed]
[ Parent | on HN ]
Re:Claude 3.7 Sonnet and Claude Code by attentive on Tuesday February 25, 2025 @05:29PM (#43178388)
given 3.5 and 3.7 cost the same, it doesn't make sense to use 3.5 here.

I'd like to see that benchmark, but R1 + 3.7 should be cheaper than 3.7T + 3.7

[ Parent | on HN ]
Re:Claude 3.7 Sonnet and Claude Code by SparkyMcUnicorn on Tuesday February 25, 2025 @05:49PM (#43178578)
The reason 3.5 (as the editor) makes more sense to me is the edit format success rate (99.6% vs 3.7's 93.3%).

Flash 2.0 got 100% on the edit format, and it's extremely cheap, so I'm pretty curious how that would score.

[ Parent | on HN ]