I'm curious about why there's been an 80% drop in costs for the O3 API. Is this due to some tech advancement or a strategy to boost sales? Would love to hear insights from developers or internal teams on this.
4 Answers
It seems like a classic case of competition heating up. Providers have to lower prices to keep customers from jumping ship to rivals. It's all about creating a solid developer ecosystem right now. If they don’t act, someone else will!
People are starting to get a bit conspiracy-minded about it, but it could be simply that they’ve found a way to use the same model at a lower cost. They probably invested heavily in GPUs to make this happen!
Some believe it’s a bit of misdirection, like they’re pulling a fast one to hide something else, but honestly, that’s just speculation. This kind of drastic price change isn’t something typically hidden from users.
I think they might be testing out new hardware. OpenAI has brought on many electrical engineers to build custom chips for their own model inference. If they can manage to do this cheaper than NVIDIA's hardware, it could position them really well in the market.
Yeah, these rumors pop up all the time. It’s not like they’ve suddenly decided to cut costs dramatically just to confuse us!