O3 is really good. I haven’t had the same results with o4 unfortunately
It's one of the most unfortunate naming conventions
OpenAI dropped the price of o3 by 80%
351–360 of 518 posts
Re: OpenAI dropped the price of o3 by 80%
#352Only at HN can the reaction to an 80% price drop be a wall of criticism.
"80% price drop" is just a title. The wall of criticism is for the fine-print.
Re: OpenAI dropped the price of o3 by 80%
#353Earlier quoted context omitted.
> OpenAI routes API requests to servers that recently processed the same prompt, My mind immediately goes to rowhammer for some reason. At the very least this opens up the possibility of some targeted denial of service
Later they mention that they have some kind of rate limiting because if over ~15 requests are being processed per minute, the request will be sent to a different server. I guess you could deny cache usage but I'm not sure what isolation they have between different callers so maybe even that won't work.
So the doc mentions you can influence the cache key by passing an optional user parameter. It’s unclear from the doc whether the user parameter is validated or if you can just provide an arbitrary string.
Re: OpenAI dropped the price of o3 by 80%
#354Earlier quoted context omitted.
[flagged]
Bit strange to see such new accounts already priming everyone to think this is acceptable and normal, which it isn't, hence OA having started doing this months ago and still not a single other provider has it, despite offering just as powerful models. It's not like Google and Anthropic have launched their own Worldcoin either.
Especially what does it have to do with that "these models are getting more powerful"?
Re: OpenAI dropped the price of o3 by 80%
#355Earlier quoted context omitted.
I suspect their data collection might not be legal in the EU. https://withpersona.com/legal/privacy-policy To me it looks like an extremely aggressive data pump.
There are stories about e.g. Hetzner requiring all sorts of data from people who want to open/verify accounts so perhaps not. Might just be an anti “money laundering” thing. Especially if the credit card company ends up refunding everything..
What stories? Can you back up that claim with some sources please?
Re: OpenAI dropped the price of o3 by 80%
#356Earlier quoted context omitted.
o3 probably used to have a HUGE profit margin on inference, so I'd say it's unclear how much optimo was done;
I find it pretty plausible they got an 80% speedup just by making optimized kernels for everything. Even when GPUs say they're being 100% utilized, there are so many improvements to be made, like: - Carefully interleaving shared memory loading with computation, and the whole kernel with global memory loading. - Warp shuffling for softmax. - Avoiding memory access conflicts in matrix multiplication. I'm sure the guys…
yes I agree that it is very plausible. But it's just unclear whether it is more of a business decision or a real downstream effect of engineering optimizations (which I assume are happening everyday at OA)
Re: OpenAI dropped the price of o3 by 80%
#357Earlier quoted context omitted.
I got the opposite experience. Not with Opus (too expensive), but with Sonnet. I got things done way more efficiently when using Sonnet with Roo than with Claude Code.
same. i ran a few tests ($100 worth of api calls) with opus 4 and didn’t see any difference compared to sonnet 4 other than the price. also no idea why he thinks roo is handicapped when claude code nerfs the thinking output and requires typing “think”/think hard/think harder/ultrathink just to expand the max thinking tokens.. which on ultrathink only sets it at 32k… when the max in roo is 51200 and it’s just a settin…
I think I could share a trick that could help:
From my experience (so not an ultimate truth) Claude is not so great at taking the decision for planning by its own: it dives immediately into coding.
If you ask it to think step-by-step it still doesn’t do it but Gemini 2.5 Pro is good at that planning but terrible at actual coding.
So you can use Gemini as planner and Claude as programmer and you get something decent on RooCode.
This “think wisely” that you have to repeat 10x in the prompt is absolutely true
Re: OpenAI dropped the price of o3 by 80%
#358I've been turned off with OpenAI and have been actively avoiding using any of their models for a while, luckily this is easy to do given the quality of Sonnet 4 / Gemini Pro 2.5.
Although I've always wondered how OpenAI could get away with o3's astronomical pricing, what does o3 do better than any other model to justify their premium cost?
Re: OpenAI dropped the price of o3 by 80%
#359Earlier quoted context omitted.
Very, very slowly. OpenAI vs Anthropic on Google Trends https://trends.google.com/trends/explore?date=today%203-m&q=... ChatGPT vs Claude on Google Trends https://trends.google.com/trends/explore?date=today%203-m&q=...
I wonder how much of this is brand name? Like Kleenex. Non-tech people might not search for LLM, generative AI, etc. ChatGPT may just be what people have heard of. I’m assuming OpenAI has a large advantage over Anthropic, and the name helps, but I bet the name is exaggerating the difference here a bit. Not everyone buys Kleenex branded Kleenex.
You are not going to find Claude when googling for ChatGPT
Re: OpenAI dropped the price of o3 by 80%
#360Google has been catching up. Funny how fast this space is evolving. Just a few months ago, it was all about DeepSeek.
If the competition boils down to who has access to the largest amount of high quality data, it's hard to see how anyone but Google could win in the end: through Google Books they have scans of tens of millions of books, and published books are the highest quality texts there are.