OpenAI just raised $122B. Frontier inference pricing hasn't moved in 9 weeks

OpenAI just closed the largest venture round in history at $852B valuation. Record capital, record confidence in AI's future.
But here's what's interesting from a market pricing perspective. Frontier model pricing has been completely flat for 9 consecutive weeks. The benchmark sits at $0.005714 per 1K input tokens across top tier flagship models globally.
At the same time the spread between frontier and budget models is 7.1x. That's a significant gap that's been holding steady.
So the question the market is now asking is which way does this go from here. Does record capital give frontier labs room to hold pricing while budget models keep improving? Or does the competitive pressure eventually compress the premium?
For teams building on top of inference at scale this dynamic matters a lot. The model selection decision isn't just a capability question anymore, it's a cost strategy question.
Curious what others think. Are you seeing this pressure in your own stack decisions?
We publish weekly inference pricing intelligence at a7om.com if you want the underlying data.