VRAMGlass contains affiliate links. As an eBay Partner, we may be compensated if you make a purchase. As an Amazon Associate I earn from qualifying purchases.We may earn from eBay and Amazon links.Learn more

GPUs tracked64Price history73 daysPrice points (24h)30,814 pointsBiggest 24h dropCMP 170HX 64GB-10.8%Biggest 24h riseRadeon PRO W7800+33.6%Updated
$ USD

GeForce RTX 5060 Price, 8GB VRAM & Local LLMs

NVIDIA's RTX 5060 (2025) is an entry-level gaming card with 8GB of VRAM that runs Qwen3.5 9B at 4-bit at about 54 tokens/s, but a 12B at 4-bit won't fit. It comes only in an 8GB version, so a listing that says “5060 16G” is most likely a 5060 Ti.

eBay

GeForce RTX 5060 price history

Current$494-0.20%
eBay
Price updated
$778$673$569$465$360
Low $420High $718
Nov 9Nov 11Nov 14
Latest price range
$420 – $718
24-hour change
-0.20%
7-day change
-1.00%
30-day change
-1.00%
View 61 listings on 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.

GeForce RTX 5060 VRAM & full specs

  • Memory8 GB · GDDR7
  • Bandwidth448 GB/s
  • FP16 / BF1676.8 TFLOPS
  • FP8 / INT8153.5 / 153.5 TOPS
Bus width
128-bit
Architecture
Blackwell (GB206)
Cores
3,840
Tensor cores
120
Ecosystem
CUDA
Interface
PCIe 5.0 x8
NVLink
Not supported
Power
145 W
Power connector
1× 8-pin / 16-pin 12V-2x6
Type
Consumer
Launch date
May 19, 2025
MSRP
$299

Best GeForce RTX 5060 deals

Powered by VRAMGlass real-time price tracking — handpicking the best-value hardware deals across the web

Picked from

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.

  1. UsedMSI G50608V3C GeForce RTX 5060 8GB Ventus 3X OC Graphics CardTop Rated99.7% positive (151.5K+ ratings)99.7% positive (151.5K+)$440Buy on eBay
  2. UsedGIGABYTE GeForce RTX 5060 WINDFORCE OC 8G GDDR7 Graphics Card GV-N5060WF2OC-8GDTop Rated100% positive (437 ratings)100% positive (437)$440Buy on eBay
  3. UsedPNY GeForce RTX 5060 8GB RAM - Graphics Card - ExcellentTop Rated99.6% positive (192K+ ratings)99.6% positive (192K+)$450Buy on eBay
  4. UsedZotac GeForce RTX 5060 Twin Edge OC 8GB GDDR7 Graphics Card 288-IN780-20076Top Rated99.6% positive (6,780 ratings)99.6% positive (6,780)$450Buy on eBay
  5. UsedZotac Gaming GeForce RTX 5060 Twin Edge OC 8GB GDDR7 128BIT Graphics Card99.7% positive (152.2K+ ratings)99.7% positive (152.2K+)$450Buy on eBay
  6. UsedGIGABYTE GeForce RTX 5060 WINDFORCE OC 8G 8GB GDDR7 Graphics Card GPU99.8% positive (3,701 ratings)99.8% positive (3,701)$450Buy on eBay
See more on eBay
Picked from

As an Amazon Associate I earn from qualifying purchases. The price you pay stays the same.

  1. NewASUS Dual NVIDIA GeForce RTX 5060 8GB GDDR7 OC Edition (PCIe 5.0, 8GB GDDR7, DLSS 4, HDMI 2.1b, DisplayPort 2.1b, 2.5-Slot Design, Axial-tech Fan Design, 0dB Technology), 3 Year Warranty$470Buy on Amazon
  2. NewASUS Dual GeForce RTX 5060 8GB GDDR7 White OC Edition Gaming Graphics Card$510Buy on Amazon
  3. NewGIGABYTE GeForce RTX 5060 WINDFORCE MAX OC 8G Graphics Card, Cooling System, 8GB 128-bit GDDR7, PCIe 5.0, by NVIDIA, DisplayPort & HDMI - Video Output Interface, GV-N5060WF2MAX OC-8GD Video Card4.5 stars (32 ratings)4.5 stars (32)$530Buy on Amazon
See more on Amazon

What LLMs can a single GeForce RTX 5060 run?

A single RTX 5060 suits models up to 9B, best at 4-bit, where Qwen3.5 9B gets up to 54K of context. A 12B fits only at 2-bit, at about 64 tokens/s; 30B-class MoE models have to keep their experts in system RAM, so speed depends on RAM bandwidth.

ModelSizeRuns?Speed t/sMax context
Gemma 4 E4B8B5 GBRuns61.8128K
Qwen3.5 9B9.7B5.7 GBRuns54.454K
Gemma 4 12B12B7.1 GBToo big——
Gemma 4 26B A4B25.2B (3.8B active)16.9 GBNeeds RAM offload35.1221K
GLM 4.7 Flash30B (3B active)18.3 GBNeeds RAM offload34.7102K
Qwen3.6 35B A3B36B (3B active)22.1 GBNeeds RAM offload47.4250K

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

  1. Small models for everyday chat and translation

    Gemma 4 E4B at 4-bit takes only 5 GB, leaving room for the full 128K of context

  2. You want to get 30B-class MoE models running

    Qwen3.6 35B A3B at 2-bit, with its experts in system RAM, runs at about 72 tokens/s, faster than a 12B running entirely on the card

  3. Small case or older power supply

    The card draws 145W, NVIDIA recommends a 550W power supply for the system, and one 8-pin connector is enough

Not a good fit

  1. You want a 12B at 4-bit

    Gemma 4 12B at 4-bit has 7.1 GB of weights, and with context and runtime overhead added it won't fit in 8GB of VRAM

  2. You want 27B-class dense models

    Even Qwen3.8 27B at 2-bit needs 9.8 GB, which an 8GB card can't hold, and putting part of it in system RAM doesn't help either

Before you buy

  1. Almost as fast as the 5060 Ti 8GB for running models

    Both cards have 8GB of VRAM and 448 GB/s of bandwidth, and generation speed follows bandwidth, so they're nearly identical. The 5060 has a sixth fewer cores, which only shows when you feed it a long document or a long conversation at once: it processes the prompt a bit slower. For the other card's numbers, see RTX 5060 Ti 8GB.

  2. The laptop 5060 is over 10% slower

    The RTX 5060 Laptop GPU also has 8GB, but only 3328 cores and 384 GB/s of bandwidth, so it generates about 14% slower, and its power is held back by the laptop it's built into. If you see a “5060 laptop” or a gaming laptop, don't estimate from the numbers on this page.

  3. Install recent software

    The 50 series is recognized only by software built with CUDA 12.8 or later. For llama.cpp and Ollama, just install the latest version; PyTorch has matching prebuilt packages starting with 2.7. If building vLLM yourself or using an old Docker image gives an error that sm_120 is unsupported, your version is too old.

Compared with similar GPUs

GeForce RTX 5060 vs GeForce RTX 5060 Ti 8GB, GeForce RTX 4060, and GeForce RTX 3060 12GB: price, VRAM, bandwidth, compute and local LLM speed compared.

SpecGeForce RTX 5060GeForce RTX 5060 Ti 8GBGeForce RTX 4060GeForce RTX 3060 12GB
Memory
8 GB
8 GBSame
8 GBSame
12 GB50% more
Bandwidth
448 GB/s
448 GB/sSame
272 GB/s39% lower
360 GB/s20% lower
BF16 / FP16 compute
77 TFLOPS
95 TFLOPS24% higher
61 TFLOPS21% lower
51 TFLOPS34% lower
INT8 compute
154 TOPS
190 TOPS24% higher
121 TOPS21% lower
102 TOPS34% lower
Qwen3.8 27B speed (2-bit)
Not supported
Not supported
Not supported
26.4 t/s
Largest dense model (4-bit)
10B
10BSame
10BSame
16B60% more
Power
145 W
180 W24% higher
115 W21% lower
170 W17% higher
Interface
PCIe 5.0 x8
PCIe 5.0 x8
PCIe 4.0 x8
PCIe 4.0 x16
Released
2025-05
2025-04
2023-06
2021-02
eBay median
$494Buy GeForce RTX 5060
$565Buy GeForce RTX 5060 Ti 8GB
$420Buy GeForce RTX 4060
$410Buy GeForce RTX 3060 12GB

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.

Other markets

Amazon.jp · Japan
$558
23 listings
View
Xianyu · Mainland China
$476
17 listings
View