I feel like that is part of their cloud strategy. If your company wants to pump a huge amount of data through one of these you will pay a premium in network costs. Their sales people will use that as a lever for why you should migrate some or all of your fleet to their cloud.

▲

jiggawatts 5 days ago | parent [-]

A few gigabytes of text is practically free to transfer even over the most exorbitant egress fee networks, but would cost “get finance approval” amounts of money to process even through a cheaper model.

▲

lbhdc 4 days ago | parent [-]

It sounds like you already know what sales peoples incentives are. They don't care about the tiny players who wanna use tiny slices. I was referring to people who are trying to push PB through these. GCPs policies make a lot of sense if they are trying to get major players to switch their compute/data host to reduce overall costs.

▲

jiggawatts 4 days ago | parent [-]

The cost ratio is the same.

▲

lbhdc 4 days ago | parent [-]

It isn't. You don't pay "msrp" at scale :)

	▲	jiggawatts 2 days ago \| parent [-]
		You're off by orders of magnitude. A million tokens is about 5 MB of text and costs $0.20 to process in something like Gemini 3 Flash. Hence, a terabyte of text would cost about $42,000 to run through a pareto-frontier "cheap" model. The most expensive cloud egress fee I could quickly find is $185 per terabyte (Azure South America Internet egress). Hence, AI processing is 200x as expensive as bandwidth, or put another way, even a long-distance international egress at "exorbitant cloud retail pricing" is a mere 0.5% of the cost. Petabytes, exabytes, etc... just adds digits to both the token cost and bandwidth cost in sync and won't significantly shift the ratio. If anything, bandwidth costs will go down and AI costs go up because: output tokens, smarter models, retries, multiple questions for the same data, etc...