GeForce RTX 5080 Price & Used Prices: Which Local LLMs Can It Run?
Price alert
We'll email you when the eBay (US) price reaches your target.
NVIDIA's RTX 5080 (2025) is a gaming card with only 16GB: 12B at 4-bit runs about 88 tokens/s, but 27B won't fit. It draws 360W; use an 850W PSU.
eBay
GeForce RTX 5080 price history
eBay
As an eBay Partner, VRAMGlass may be compensated if you buy through the eBay links on this page. The price you pay stays the same.
Full GeForce RTX 5080 specifications
- Memory16 GB · GDDR7
- Bandwidth960 GB/s
- FP16 / BF16225.1 TFLOPS
- FP8 / INT8450.2 / 450.2 TOPS
- Bus width
- 256-bit
- Architecture
- Blackwell (GB203)
- Cores
- 10,752
- Tensor cores
- 336
- Ecosystem
- CUDA
- Interface
- PCIe 5.0 x16
- NVLink
- Not supported
- Power
- 360 W
- Power connector
- 16-pin 12V-2x6
- Slots
- 2
- Length
- 304 mm
- Type
- Consumer
- Launch date
- Jan 30, 2025
- MSRP
- $999
Best GeForce RTX 5080 deals on eBay
Hand-picked deals from trusted eBay sellers. Shop with confidence.
As an eBay Partner, VRAMGlass may be compensated if you buy through the eBay links on this page. The price you pay stays the same.
UsedMSI Nvidia GeForce RTX 5080 16G GAMING TRIO OC Graphic Card PCIE Gen 5$1,650Buy on eBay
UsedPNY ARGB EPIC-X Overclocked Triple Fan GeForce RTX 5080 16GB - GPU - ExcellentTop Rated$1,660Buy on eBay
UsedPNY Overclocked Triple Fan GeForce RTX 5080 16GB - Graphics Card - ExcellentTop Rated$1,700Buy on eBay
NewALIENWARE Nvidia RTX 5080 16GB GDDR7 0CRPP5 GPU NEW ✅ 🇺🇸Top Rated$1,700+ shippingBuy on eBay
UsedGigabyte GeForce RTX 5080 AERO OC 16GB GDDR7 Graphics Card GV-N5080AERO OC-16GDTop Rated$1,700Buy on eBay
UsedNVIDIA GeForce RTX 5080 16GB Dell OEM$1,799+ shippingBuy on eBay
What LLMs can a single GeForce RTX 5080 run?
A single RTX 5080's best big model is Qwen3.6 35B A3B at 2-bit, about 161 tokens/s; at 4-bit with system RAM it falls to 40-50. Up to 12B, use 4-bit.
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
 Gemma 4 E4B8B | 3.8 GB | Runs | 150 | 128K |
 Qwen3.5 9B9.7B | 4.1 GB | Runs | 139.9 | 256K |
| Gemma 4 12B12B | 4.7 GB | Runs | 123.5 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 10.5 GB | Runs | 140.8 | 256K |
| Qwen3.6 27B27.8B | 11.8 GB | Runs | 56.2 | 64K |
| Qwen3.8 27B27.8B | 9.8 GB | Runs | 66.3 | 93K |
 GLM 4.7 Flash30B (3B active) | 11.9 GB | Runs | 151.2 | 75K |
| Gemma 4 31B30.7B | 11.8 GB | Runs | 54.2 | 41K |
| Qwen3.6 35B A3B36B (3B active) | 12.3 GB | Runs | 161.1 | 181K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 5 GB | Runs | 120.5 | 128K |
| Qwen3.5 9B9.7B | 5.7 GB | Runs | 107.2 | 256K |
| Gemma 4 12B12B | 7.1 GB | Runs | 88 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 16.9 GB | Needs RAM offload | 38.9 | 256K |
| Qwen3.6 27B27.8B | 16.8 GB | Too big | — | — |
| Qwen3.8 27B27.8B | 16.5 GB | Too big | — | — |
| GLM 4.7 Flash30B (3B active) | 18.3 GB | Needs RAM offload | 36.6 | 198K |
| Gemma 4 31B30.7B | 18.3 GB | Too big | — | — |
| Qwen3.6 35B A3B36B (3B active) | 22.1 GB | Needs RAM offload | 52.3 | 256K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 8.2 GB | Runs | 79.1 | 128K |
| Qwen3.5 9B9.7B | 9.5 GB | Runs | 69 | 196K |
| Gemma 4 12B12B | 12.7 GB | Runs | 52.7 | 183K |
| Gemma 4 26B A4B25.2B (3.8B active) | 26.9 GB | Needs RAM offload | 26.3 | 256K |
| Qwen3.6 27B27.8B | 28.6 GB | Too big | — | — |
| Qwen3.8 27B27.8B | 29 GB | Too big | — | — |
| GLM 4.7 Flash30B (3B active) | 31.8 GB | Needs RAM offload | 22.8 | 198K |
| Gemma 4 31B30.7B | 32.6 GB | Too big | — | — |
| Qwen3.6 35B A3B36B (3B active) | 36.9 GB | Needs RAM offload | 34.8 | 256K |
RunsFits fully in VRAM; speed is estimated from VRAM bandwidth.
Needs RAM offloadMoE only: expert weights sit in system RAM, so speed is estimated from 70 GB/s RAM bandwidth.
Too bigWon’t fit on one card at this quantization, even with RAM offload.
Full calculation rules are on the Methodology page.
Who it suits
Good fit
- The fastest speeds on models that fit in 16GB
Gemma 4 12B at 4-bit runs at about 88 tokens/s, and Qwen3.6 35B A3B at 2-bit at about 161
- Regularly feeding in a whole codebase or long documents at once
It has the most compute of any 16GB card on our site, about 30% more than the 5070 Ti
- You want a high-end card in two slots
The Founders Edition is dual-slot and 304 mm long
Not a good fit
- Fitting the next tier of models
It still has 16GB, so Qwen3.8 27B at 4-bit (16.5 GB) still doesn't fit
- A tight power supply
The card draws 360W, and NVIDIA recommends an 850W system power supply, one step up from the 5070 Ti
Before you buy
Dual-slot Founders Edition, but bring enough power
The Founders Edition is dual-slot and 304 mm long, noticeably thinner than the three-slot 4080 Founders Edition. But the card draws 360W, and NVIDIA recommends an 850W power supply. With the adapter, you need three 8-pin cables from the power supply; with a native 12V-2x6 cable, it should be rated for 450W or more.
Check what the AI TOPS figure measures
The 1801 AI TOPS NVIDIA advertises is an FP4 figure with sparsity. Measured the way our comparisons do (BF16, without sparsity), it is only about 225 TFLOPS. When comparing compute with other cards, first make sure the figures are measured the same way.
Compared with similar GPUs
GeForce RTX 5080 vs GeForce RTX 5070 Ti, GeForce RTX 4080, and GeForce RTX 4090: price, VRAM, bandwidth, compute and local LLM speed compared.
| Spec | GeForce RTX 5080 | GeForce RTX 5070 Ti | GeForce RTX 4080 | GeForce RTX 4090 |
|---|---|---|---|---|
| Memory | 16 GB | 16 GBSame | 16 GBSame | 24 GB50% more |
| Bandwidth | 960 GB/s | 896 GB/s7% lower | 717 GB/s25% lower | 1,008 GB/s5% higher |
| BF16 / FP16 compute | 225 TFLOPS | 176 TFLOPS22% lower | 195 TFLOPS13% lower | 330 TFLOPS47% higher |
| INT8 compute | 450 TOPS | 352 TOPS22% lower | 390 TOPS13% lower | 661 TOPS47% higher |
| Qwen3.8 27B speed (2-bit) | 66.3 t/s | 62.3 t/s6% slower | 50.7 t/s24% slower | 69.3 t/s5% faster |
| Largest dense model (4-bit) | 23B | 23BSame | 23BSame | 37B61% more |
| Power | 360 W | 300 W17% lower | 320 W11% lower | 450 W25% higher |
| Interface | PCIe 5.0 x16 | PCIe 5.0 x16 | PCIe 4.0 x16 | PCIe 4.0 x16 |
| Released | 2025-01 | 2025-02 | 2022-11 | 2022-10 |
| eBay median |
As an eBay Partner, VRAMGlass may be compensated if you buy through the eBay links on this page. The price you pay stays the same.
GeForce RTX 5080 rental price per hour
On-demand rental of the GeForce RTX 5080 starts at $0.39 per hour (RunPod), about $95/month at 8 hours a day. Buying one (eBay median $2,074) pays off in about 4.4 years, 2.2 years or 9 months at 4, 8 or 24 hours a day.
| Option | Price | 4 h/day | 8 h/day | 24 h/day | View |
|---|---|---|---|---|---|
 RunPodRentUpdated 47 min. ago | $0.39/hr | $47/mo | $95/mo | $285/mo | Rent |
 Vast.aiRentUpdated 47 min. ago | $0.67/hr | $81/mo | $163/mo | $488/mo | Rent |
 eBayUpdated 6 hr. ago | $2,074 | Payback ~4.4 yrsPower $8/mo | Payback ~2.2 yrsPower $16/mo | Payback ~9 moPower $47/mo | Buy |
Break-even uses the US residential average ($0.18/kWh) and 360 W rated power, excluding resale value and the rest of the build.
Rental links are referral links: we may receive a commission or credit if you sign up through them. As an eBay Partner, VRAMGlass may be compensated if you buy through the eBay links on this page. The price you pay stays the same.
Amazon.jp
Xianyu