I've been using the O3 model pretty regularly, and I noticed a weird pattern. No matter how complicated my prompts are, it always finishes in exactly 12 seconds. I've seen other models take anywhere from 30 seconds to over a minute, but O3 seems to stick to its 12-second routine. The only time it deviated was when I gave it a challenging image with cut-off letters, and it thought for about 6 minutes, but ironically, it got the answer wrong anyway. I'm on the plus plan and I'm curious if there's a way to get O3 to think longer and provide better-quality responses without having to switch to the pro plan, which seems pricey.
1 Answer
You're not alone! I've noticed the same thing with O3. I ended up switching to Gemini 2.5 Pro because I found O3's responses to be pretty limiting. Maybe give that a shot if you're looking for more depth?
How's Gemini Pro compared to ChatGPT? I'm considering it for web development, but I've really struggled with O3.