←back to thread

OpenAI o3 and o4-mini

(openai.com)
555 points maheshrijal | 4 comments | 16 Apr 25 17:01 UTC | HN request time: 0.709s | source
Show context
jdross ◴[16 Apr 25 17:12 UTC] No.43707849[source]▶
The pace of notable releases across the industry right now is unlike any time I remember since I started doing this in the early 2000's. And it feels like it's accelerating
replies(3): >>43707964 #>>43708571 #>>43712041 #
qoez ◴[16 Apr 25 18:08 UTC] No.43708571[source]▶
Lots of releases but very little actual performance increases
replies(1): >>43708812 #
1. int_19h ◴[16 Apr 25 18:34 UTC] No.43708812[source]▶
Sonnet and Gemini saw fairly substantial perf increases recenly
replies(1): >>43709040 #
2. mchusma ◴[16 Apr 25 18:55 UTC] No.43709040[source]▶
Love Sonnet but 3.7 is not obviously an improvement over 3.5 in my real world usage. Gemini 2.5 pro is great, has replaced most others for me (Grok I use for things that require realtime answers)
replies(2): >>43710308 #>>43711956 #
3. BriggyDwiggs42 ◴[16 Apr 25 20:57 UTC] No.43710308[source]▶
It does a lot better on philosophy questions.
4. int_19h ◴[17 Apr 25 00:55 UTC] No.43711956[source]▶
Are you comparing it with or without thinking? I'd say it's a fairly big improvement in long thinking mode.