GeForce RTX 4060 Ti 16GB Price, VRAM & Local LLMs
Price alert
We'll email you when the eBay (US) price reaches your target.
The RTX 4060 Ti 16GB (2023) pairs entry-level bandwidth, just 288 GB/s, with enough VRAM to run 30B-class MoE at 2-bit at about 103 tokens/s. Make sure a listing is the 16GB card, not the 8GB one.
eBay
GeForce RTX 4060 Ti 16GB price history
Current$7490.00%
eBay
Price updated
$913$803$693$582$472
Low $669High $830
Sep 23Oct 19Nov 14
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 4060 Ti 16GB VRAM & full specs
- Memory16 GB · GDDR6
- Bandwidth288 GB/s
- FP16 / BF1688.3 TFLOPS
- FP8 / INT8176.5 / 176.5 TOPS
- Bus width
- 128-bit
- Architecture
- Ada Lovelace (AD106)
- Cores
- 4,352
- Tensor cores
- 136
- Ecosystem
- CUDA
- Interface
- PCIe 4.0 x8
- NVLink
- Not supported
- Power
- 165 W
- Power connector
- 1× 8-pin / 16-pin 12VHPWR
- Type
- Consumer
- Launch date
- Jul 18, 2023
- MSRP
- $499
Best GeForce RTX 4060 Ti 16GB 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.
UsedMSI GeForce RTX 4060 Ti Ventus 3X 16GB GDDR6 OC Graphics Card PCIe 4 + WARRANTYTop Rated$749$999-25%Buy on eBay
UsedASUS Dual RTX 4060 Ti 16GB Advanced Edition Graphics Card DUAL-RTX4060TI-A16GTop Rated$760Buy on eBay
UsedAsus GeForce RTX 4060 Ti 16GB GDDR6 Graphics Card AD106-350Top Rated$800Buy on eBay
UsedASUS ProArt RTX 4060 Ti OC 16GB GDDR6 - 1 Hour FurMark Tested$800+ shippingBuy on eBay
UsedASUS GeForce RTX 4060 Ti 16GB Dual EVO OC GPU | 1yr Warranty, Fast Ship!Top Rated$1,050Buy on eBay
UsedGigabyte GeForce RTX 4060 Ti 16GB AERO OC GPU | 1yr Warranty, Fast Ship!Top Rated$1,050Buy on eBay
Picked from
As an Amazon Associate I earn from qualifying purchases. The price you pay stays the same.
UsedGeForce RTX 4060 Ti Ventus 3X 16G OC Graphics Card -NVIDIA RTX 4060 Ti, 16GB GDDR6 Memory, 18Gbps, PCIe 4.0, DLSS3 (Renewed)$764$950-20%Buy on Amazon- Newmsi GeForce RTX 4060 Ti Ventus 3X 16G OC Graphics Card -NVIDIA RTX 4060 Ti, 16GB GDDR6 Memory, 18Gbps, PCIe 4.0, DLSS3$950Buy on Amazon
UsedGaming GeForce RTX 4060 Ti 16GB AMP DLSS 3 16GB GDDR6 128-bit 18 Gbps PCIE 4.0 Compact Gaming Graphics Card, IceStorm 2.0 Advanced Cooling, Spectra RGB Lighting, ZT-D40620F-10M (Renewed)$1,070Buy on Amazon
What LLMs can a single GeForce RTX 4060 Ti 16GB run?
One 4060 Ti 16GB fits only up to 12B at 4-bit, Gemma 4 12B under 30 tokens/s. Its best run is Qwen3.6 35B A3B at 2-bit, all on the card, about 100.
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
 Gemma 4 E4B8B | 3.8 GB | Runs | 52.7 | 128K |
 Qwen3.5 9B9.7B | 4.1 GB | Runs | 48.6 | 256K |
| Gemma 4 12B12B | 4.7 GB | Runs | 42.1 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 10.5 GB | Runs | 78.7 | 256K |
| Qwen3.6 27B27.8B | 11.8 GB | Runs | 17.8 | 64K |
| Qwen3.8 27B27.8B | 9.8 GB | Runs | 21.3 | 93K |
 GLM 4.7 Flash30B (3B active) | 11.9 GB | Runs | 90.3 | 75K |
| Gemma 4 31B30.7B | 11.8 GB | Runs | 17.2 | 41K |
| Qwen3.6 35B A3B36B (3B active) | 12.3 GB | Runs | 102.9 | 181K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 5 GB | Runs | 41 | 128K |
| Qwen3.5 9B9.7B | 5.7 GB | Runs | 36 | 256K |
| Gemma 4 12B12B | 7.1 GB | Runs | 28.9 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 16.9 GB | Needs RAM offload | 31.9 | 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 | 33 | 198K |
| Gemma 4 31B30.7B | 18.3 GB | Too big | — | — |
| Qwen3.6 35B A3B36B (3B active) | 22.1 GB | Needs RAM offload | 43.2 | 256K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 8.2 GB | Runs | 25.7 | 128K |
| Qwen3.5 9B9.7B | 9.5 GB | Runs | 22.2 | 196K |
| Gemma 4 12B12B | 12.7 GB | Runs | 16.7 | 183K |
| Gemma 4 26B A4B25.2B (3.8B active) | 26.9 GB | Needs RAM offload | 21.5 | 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 | 20.6 | 198K |
| Gemma 4 31B30.7B | 32.6 GB | Too big | — | — |
| Qwen3.6 35B A3B36B (3B active) | 36.9 GB | Needs RAM offload | 28.3 | 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.
What can the GeForce RTX 4060 Ti 16GB run with Strata?
With Strata, the GeForce RTX 4060 Ti 16GB can run models far larger than its VRAM. Strata keeps most of the model in RAM and on the SSD and only the most-used experts on the GPU. Benchmarks come from Strata’s author and community, with multi-GPU setups marked by card count and today’s prices added by VRAMGlass.
Where Strata puts the model
Qwen3.8 Flash Next Q2_0 · 66.4 GB- VRAM 16 GB: the parts every token uses, plus the most-used experts
- RAM about 40 GB: all the experts; the CPU computes the ones the GPU doesn’t hold
- SSD 28.8 GB: the lookup table, a few rows read per token
1.75 bpw · 239 GBMeasured
Qwen3.8 Flash Next
IQ2_XS · 68 GBMeasuredQwen3.8 Flash Next
UD-IQ4_XS · 94 GBMeasured- 2026-10-05
0.1.39- Pop!_OS 24.04
- @fioruccione
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.
Who it suits
Good fit
- You want 16GB with less heat
The card draws 165W, and NVIDIA recommends a 550W power supply for the system
- Running 30B-class MoE models at 2-bit
Qwen3.6 35B A3B takes only 12.3 GB and runs entirely on the card at about 103 tokens/s with 181K of context
- You want a 12B at 8-bit
Gemma 4 12B at 8-bit (12.7 GB) fits with up to 183K of context, at about 17 tokens/s
Not a good fit
- Generation speed matters to you
Its bandwidth is the lowest of any NVIDIA 16GB card on this site, and Gemma 4 12B at 4-bit manages only about 29 tokens/s
- You want dense models of 27B and up
They don't fit at 4-bit, so you're limited to 2-bit, where Qwen3.8 27B runs at about 21 tokens/s
Before you buy
The 16GB version comes only from partner brands
The 4060 Ti comes in 8GB and 16GB versions with the same core specs, and NVIDIA made a Founders Edition only for the 8GB version. Used listings often mix the two up, so before ordering, ask the seller for a GPU-Z or nvidia-smi screenshot confirming 16GB.
Only 8 PCIe lanes
It's wired for only 8 PCIe 4.0 lanes, so on an older PCIe 3.0 motherboard it drops to 3.0 x8, a quarter of the bandwidth of full-speed 4.0 x16. For single-card inference, PCIe is mostly busy only while loading the model, so the impact is small. If you want to add cards for multi-GPU later, we recommend at least x8 per card, and on an older platform this card sits right at that minimum.
A big cache doesn't make up for bandwidth
The 4060 Ti has 32MB of L2 cache, which helps in games but not with running models: every generated token requires reading the full set of weights from VRAM, several GB that the cache can't hold, so speed is still set by the 288 GB/s memory bandwidth.
Compared with similar GPUs
GeForce RTX 4060 Ti 16GB vs GeForce RTX 5060 Ti 16GB, GeForce RTX 3060 12GB, and Radeon RX 7800 XT: price, VRAM, bandwidth, compute and local LLM speed compared.
| Spec | GeForce RTX 4060 Ti 16GB | GeForce RTX 5060 Ti 16GB | GeForce RTX 3060 12GB | Radeon RX 7800 XT |
|---|---|---|---|---|
| Memory | 16 GB | 16 GBSame | 12 GB25% less | 16 GBSame |
| Bandwidth | 288 GB/s | 448 GB/s56% higher | 360 GB/s25% higher | 624 GB/s117% higher |
| BF16 / FP16 compute | 88 TFLOPS | 95 TFLOPS7% higher | 51 TFLOPS42% lower | 75 TFLOPS15% lower |
| INT8 compute | 177 TOPS | 190 TOPS7% higher | 102 TOPS42% lower | 75 TOPS58% lower |
| Qwen3.8 27B speed (2-bit) | 21.3 t/s | 32.6 t/s53% faster | 26.4 t/s24% faster | 44.6 t/s109% faster |
| Largest dense model (4-bit) | 23B | 23BSame | 16B30% less | 23BSame |
| Power | 165 W | 180 W9% higher | 170 W3% higher | 263 W59% higher |
| Interface | PCIe 4.0 x8 | PCIe 5.0 x8 | PCIe 4.0 x16 | PCIe 4.0 x16 |
| Released | 2023-07 | 2025-04 | 2021-02 | 2023-09 |
| 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.
Amazon.jp
Xianyu