PerGB

NVIDIA GeForce RTX 5090

Live eBay prices, cheapest first. The best one right now is $5,000 (used).

Other sizes: 24 GB

Specs

VRAM
32 GB GDDR7
Bandwidth
1,792 GB/s
FP32
104.8 TFLOPS
FP16
209 TFLOPS
Power
575 W
Connector
PCIe 5.0
Released
2025
Launch price
$1,999
CUDA compute
12.0
Llama 3.1 8B at Q4 (est.)
185–277 tokens/s

Price right now

Live on eBay
Lowest now$5,000
Median ask$7,000
Per GB VRAM$156.25
vs launch price150% over
Live listings215

From the live listings below; shipping included.

Listings

Live on eBay
ListingConditionPriceChecked
Used$5,000or best offer−29%free shipping
New$5,050or best offer−28%shipping not listed
Used$5,750−18%free shipping
Open box$5,800−17%free shipping
Refurbished$5,999−14%free shipping
Refurbished$5,999−14%free shipping
Used$6,000−14%free shipping
Used$6,000or best offer−14%shipping not listed
Used$6,008or best offer−14%+$7.72 shipping
Used$6,050or best offer−14%+$50.00 shipping
Used$6,078or best offer−13%+$122.63 shipping
Used$6,100or best offer−13%free shipping
Used$6,108or best offer−13%+$7.72 shipping
Used$6,150−12%free shipping
New$6,198or best offer−11%free shipping
New$6,199or best offer−11%free shipping
Open box$6,200or best offer−11%free shipping
Used$6,216or best offer−11%+$15.90 shipping
Used$6,250or best offer−11%shipping not listed
Open box$6,279−10%free shipping
Used$6,298or best offer−10%+$299.00 shipping
Used$6,299−10%free shipping
Open box$6,299−10%shipping not listed
Refurbished$6,299−10%free shipping
Used$6,299or best offer−10%shipping not listed
New$6,378or best offer−9%free shipping
New$6,378or best offer−9%free shipping
Open box$6,379−9%free shipping
Used$6,382or best offer−9%+$44.09 shipping
Used$6,399or best offer−9%shipping not listed
New$6,400−9%free shipping
New$6,400or best offer−9%shipping not listed
Used$6,400−9%free shipping
Used$6,400or best offer−9%free shipping
Used$6,400−9%shipping not listed
New$6,435−8%+$35.00 shipping
New$6,438−8%free shipping
New$6,440−8%+$39.99 shipping
New$6,448or best offer−8%free shipping
Used$6,449−8%free shipping
New$6,478or best offer−7%free shipping
Used$6,499or best offer−7%free shipping
Open box$6,499−7%shipping not listed
Open box$6,499−7%free shipping
New$6,500−7%free shipping
New$6,500or best offer−7%free shipping
Open box$6,500or best offer−7%shipping not listed
Used$6,500−7%free shipping
Used$6,500or best offer−7%free shipping
Used$6,500or best offer−7%free shipping
New$6,500−7%free shipping
Used$6,500or best offer−7%free shipping
Used$6,500or best offer−7%free shipping
Open box$6,500or best offer−7%free shipping
Used$6,500−7%shipping not listed
Used$6,500or best offer−7%shipping not listed
Used$6,500or best offer−7%shipping not listed
Used$6,500or best offer−7%shipping not listed
Used$6,500or best offer−7%shipping not listed
New$6,518−7%+$33.00 shipping

Showing the 60 cheapest of 215 listings.

Auctions ending soon

No auctions ending soon right now.

Prices include shipping and were read from eBay at the time shown. They can change: check the listing before you buy.

Owner reviews

No owner reviews yet. Own one? Tell others how it runs.

From signed-in owners, checked before they're published. PerGB doesn't pay for or edit reviews.

How it compares

Live on eBay

Faster, for the least

More VRAM, for not much more

Compare: vs RTX PRO 6000 Blackwell, vs GeForce RTX 4090, vs GeForce RTX 4090 D 24 GB.

From the live listings of every GPU, shipping included; performance against the GeForce RTX 3090 (FP16 or FP32 compute and memory bandwidth). Updated with every build.

Local AI: what it runs

Model (Q4, 8K context)GeForce RTX 5090 neededCosts nowSpeed (est.)
Llama 3.1 8B1 card$5,000185–277 tokens/s
Qwen3 14B1 card$5,000100–150 tokens/s
gpt-oss 20B1 card$5,000246–369 tokens/s
Gemma 3 27B1 card$5,00054–81 tokens/s
Qwen3 32B1 card$5,00045–68 tokens/s
Llama 3.3 70B2 cards$10,05021–31 tokens/s
gpt-oss 120B3 cards$15,800174–261 tokens/s
Qwen3 235B-A22B5 cards$27,59940–60 tokens/s
DeepSeek R1 / V3 671Bmore than 8——

Cost: that many of the cheapest live listings. Speed: an estimate from the memory bandwidth for one person chatting. Try other models, sizes and contexts.

Decision models

Model (Q4, 8K context)GeForce RTX 5090 neededCosts nowPer question (est.)
Julia-10.144B1 card$5,0003.2–7.0 ms
Laya0.421B1 card$5,0005.4–12 ms
Kev-4B4B1 card$5,00013–28 ms
lev4B1 card$5,00039–84 ms
OpenJev27B1 card$5,00047–100 ms
Clef27B1 card$5,00047–100 ms
Perplexity Decider v1.227B1 card$5,00047–100 ms
mammoth-127B1 card$5,00047–100 ms

Decision models score the options you give them instead of writing text, so they're timed per question. The time is scaled from each model's reported speed on an RTX PRO 6000 by this card's compute and memory bandwidth, within about half either way. About the models.

Compare elsewhere

Search other shops for NVIDIA GeForce RTX 5090. These open the shop's own search; we don't show their prices here.