News for hackers. Stuff that matters.
Re:GPT-4.1 in the API by hmottestad on Monday April 14, 2025 @01:16PM (#43683635)
Are those with «thinking» or without?
Re:GPT-4.1 in the API by sanxiyn on Monday April 14, 2025 @02:12PM (#43684380)
Sonnet 3.7's 70% is without thinking, see https://www.anthropic.com/news/claude-3-7-sonnet
Re:GPT-4.1 in the API by energy123 on Monday April 14, 2025 @01:35PM (#43683918)
With
Re:GPT-4.1 in the API by chaos_emergent on Monday April 14, 2025 @01:46PM (#43684043)
based on their release cadence, I suspect that o4-mini will compete on price, performance, and context length with the rest of these models.
Re:GPT-4.1 in the API by hecticjeff on Monday April 14, 2025 @02:15PM (#43684423)
o4-mini, not to be confused with 4o-mini
Re:GPT-4.1 in the API by aledalgrande on Monday April 14, 2025 @09:43PM (#43688196)
The thinking tokens (even just 1024) make a massive difference in real world tasks with 3.7 in my experience