Graphics card pricing is a contentious topic. Many of us remember the days when a great mid-range GPU wouldn’t run you much beyond $300-$400, and the high-end stuff was still priced somewhat reasonably for most gamers. Halcyon days, my friends, and ripe for reminiscence. It really was better in the good old days, wasn’t it?

It’s a different world now though, one where $1,600+ RTX 4090s roam the lands. However, according to known leaker kopite7kimi, we shouldn’t be expecting a significant chunk of change on top of the MSRP for Nvidia’s upcoming top-end card. They’ve posted an arguably reassuring tweet stating that they “don’t believe there will be a significant price increase for the RTX 5090”.

Given the supposed (and truly gargantuan) specs of that particular GPU, this might come as a surprise. After all, with a rumoured 32 GB of GDDR7 across a 512-bit memory bus and potentially 21,760 CUDA cores, if those specs turn out to be accurate we’ll likely be looking at an incredible performer.

Still, there has to be a price ceiling somewhere—and I wouldn’t be surprised if Nvidia had already found it with the RTX 4090. While it may have learned that people will pay truly staggering amounts of money for similarly staggering levels of performance from its current-generation efforts, a $2,000 or similar halo card really does seem like it would be a step too far even for very well-off gamers.

That was, after all, the calculation Nvidia made while chasing the covid/GPU scarcity dollar when it released the RTX 3090 Ti. And it rowed back pretty significantly on that when the RTX 4090 was released at $1,600.

That being said, and again assuming that kopite7kimi is correct, I wonder what they classify as “significant”. After all, given the costs we’re talking about overall, an extra $100-$200 over the $1,600 MSRP of the RTX 4090 would only amount to roughly a 6% to 13% increase—and given the specs on offer, whether that’s a significant rise over the previous pricing is a matter of opinion.

That sort of increase seems about right to me, though. A $1,700 to $1,800 RTX 5090 might be much too expensive for many gamers, but RTX 4090s have continued to sell at well beyond that, and there’s likely enough of a market of well-heeled enthusiasts that’ll pay those sort of prices to jump on the most powerful Blackwell-based consumer card.

And with AMD bowing out of the high end market for this generation of GPUs? Well, it’s Nvidia’s field to dominate. With no clear competition, Nvidia has a free run to do what it likes—although the speculation about a $2,000+ card seems a little far fetched.

I’m not ruling it out, mind, but it strikes as a step too far even in the clear open skies of the high-end GPU market that Nvidia finds itself in.

As for the RTX 5080? I’d say it’s likely that Nvidia will price it around the $1,000 mark. The RTX 4080 and RTX 4080 Super released at $1,199 and $999 respectively, and the latter was rather well received for delivering almost exactly the same performance as its non-Super variant for a cheaper launch price.

But what of the RTX 5070, rumoured to be launching at CES 2025 at the same time as its faster siblings? Well, here’s where things get interesting. AMD will want to stay competitive in the mid-range with its RDNA 4 GPUs—even if it has conceded the high-end market—so there may be actual competition here.

Still, Nvidia doesn’t seem concerned in the slightest with what AMD are up to these days.

But when it comes to mid-range, price really does matter. So while I doubt the RTX 5070 will launch with the adjusted price Nvidia gave the RTX 4070 to cover off the RX 7800 XT, I wouldn’t be surprised if it was around the same as the original launch price. That’d make it $599. Add on $50 or so for, I dunno, inflation, Nvidia-tax etc., and I reckon $649 or so would be a safe bet.

Of course, I’m speculating here, and opening myself up to have this article thrown in my face if the RTX 50-series GPUs come out for truly outrageous prices and the internet has a collective meltdown.

But while graphics card pricing has seen huge rises over the past decade or so, it does feel like the top may have been found, for a while yet at least.

Andy’s great mistake? You may well have just read it, but this is where I’ll hedge my bets for now.