Radeon RX 9060 XT 16GB Price, VRAM & Local LLMs
Price alert
We'll email you when the eBay (US) price reaches your target.
AMD's RX 9060 XT 16GB (2025) is an entry-level gaming card that fits an entire 30B-class MoE model at 2-bit, running Qwen3.6 35B A3B at about 109 tokens/s. Make sure you buy the 16GB version: the 8GB version uses the same chip with half the VRAM.
eBay
Radeon RX 9060 XT 16GB 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.
Radeon RX 9060 XT 16GB VRAM & full specs
- Memory16 GB · GDDR6
- Bandwidth320 GB/s
- FP16 / BF16103 TFLOPS
- FP8 / INT8205 / 205 TOPS
- Bus width
- 128-bit
- Architecture
- RDNA 4 (Navi 44)
- Cores
- 2,048
- Tensor cores
- 64
- Ecosystem
- ROCm
- Interface
- PCIe 5.0 x16
- Power
- 160 W
- Power connector
- 1× 8-pin
- Type
- Consumer
- Launch date
- Jun 5, 2025
- MSRP
- $349
Best Radeon RX 9060 XT 16GB deals
Powered by VRAMGlass real-time price tracking — handpicking the best-value hardware deals across the web
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.
UsedPowerColor Reaper AMD Radeon RX 9060 XT 16GB Graphics Card$700Buy on eBay
UsedXFX Mercury Gaming OC Radeon RX 9060 XT 16GB - Graphics Card - ExcellentTop Rated$700Buy on eBay
NewXFX Swift AMD Radeon RX 9060 XT 16GB Graphics Card NEW SEALED$730Buy on eBay
UsedASUS Radeon RX 9060 XT Prime OC 16G Graphics Card$740+ shippingBuy on eBay
NewXFX Swift Radeon RX 9060 XT 16GB OC GDDR6 Video Graphics CardTop Rated$750Buy on eBay
UsedGIGABYTE Radeon RX 9060 XT GAMING OC 16GB Graphic Card - USED$750Buy on eBay
As an Amazon Associate I earn from qualifying purchases. The price you pay stays the same.
NewASRock Radeon RX 9060 XT Challenger Pro 16GB OC Graphics Card, RDNA 4, 3290MHz Boost, 16GB GDDR6 128-bit, PCIe 5.0, Triple Fan, 0dB Silent, LED Switch$730Buy on Amazon
NewASUS Dual Radeon RX 9060 XT 16GB GDDR6 Gaming Graphics Card$746$790-6%Buy on Amazon
NewXFX Mercury AMD Radeon RX 9060 XT OC Gaming Edition with 16GB GDDR6 HDMI 2xDP, AMD RDNA 4 RX 9060XT RX-96TMERCBA$750Buy on Amazon
What LLMs can a single Radeon RX 9060 XT 16GB run?
A single RX 9060 XT 16GB is most comfortable with models up to 12B at 4-bit, and a 12B fits at 8-bit too. 27B-class dense models work only at 2-bit, with Qwen3.8 27B at about 24 tokens/s and 93K of context.
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
 Gemma 4 E4B8B | 3.8 GB | Runs | 58.1 | 128K |
 Qwen3.5 9B9.7B | 4.1 GB | Runs | 53.6 | 256K |
| Gemma 4 12B12B | 4.7 GB | Runs | 46.5 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 10.5 GB | Runs | 84 | 256K |
| Qwen3.6 27B27.8B | 11.8 GB | Runs | 19.8 | 64K |
| Qwen3.8 27B27.8B | 9.8 GB | Runs | 23.6 | 93K |
 GLM 4.7 Flash30B (3B active) | 11.9 GB | Runs | 95.8 | 75K |
| Gemma 4 31B30.7B | 11.8 GB | Runs | 19 | 41K |
| Qwen3.6 35B A3B36B (3B active) | 12.3 GB | Runs | 108.5 | 181K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 5 GB | Runs | 45.3 | 128K |
| Qwen3.5 9B9.7B | 5.7 GB | Runs | 39.7 | 256K |
| Gemma 4 12B12B | 7.1 GB | Runs | 32 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 16.9 GB | Needs RAM offload | 32.7 | 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 | 33.4 | 198K |
| Gemma 4 31B30.7B | 18.3 GB | Too big | — | — |
| Qwen3.6 35B A3B36B (3B active) | 22.1 GB | Needs RAM offload | 44.3 | 256K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 8.2 GB | Runs | 28.5 | 128K |
| Qwen3.5 9B9.7B | 9.5 GB | Runs | 24.6 | 196K |
| Gemma 4 12B12B | 12.7 GB | Runs | 18.5 | 183K |
| Gemma 4 26B A4B25.2B (3.8B active) | 26.9 GB | Needs RAM offload | 22 | 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 | 20.9 | 198K |
| Gemma 4 31B30.7B | 32.6 GB | Too big | — | — |
| Qwen3.6 35B A3B36B (3B active) | 36.9 GB | Needs RAM offload | 29.1 | 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
- Running 30B-class MoE models entirely on the card
Qwen3.6 35B A3B at 2-bit takes only 12.3 GB, leaves room for up to 181K of context, and doesn't need system RAM
- You want a 12B at 8-bit
Gemma 4 12B at 8-bit fits entirely on the card with 183K of context
- Working with a modest power supply
The card draws 160W, AMD recommends a 450W power supply, and it takes a single 8-pin connector
Not a good fit
- You care most about generation speed
Its bandwidth is only 320 GB/s, and Gemma 4 12B at 4-bit runs at about 32 tokens/s
- You want a 27B at 4-bit
Qwen3.8 27B at 4-bit needs 16.5 GB, and with context and runtime overhead added it won't fit in 16GB of VRAM, so you have to drop to 2-bit
Before you buy
All 16 PCIe lanes
The RTX 5060 Ti in the same class uses only 8 PCIe lanes, while the 9060 XT is wired for all 16, so it still gets 3.0 x16 on an older PCIe 3.0 motherboard. Once a model is in VRAM this makes little difference; the main gain is faster loading.
Getting the software running
AMD's ROCm compatibility list includes the RX 9060 XT under Linux, and the Linux version of Ollama recognizes it directly, but this generation isn't on the Windows version's support list yet. llama.cpp runs on it with either the ROCm or the Vulkan backend.
Compared with similar GPUs
Radeon RX 9060 XT 16GB vs GeForce RTX 5060 Ti 16GB, Radeon RX 9070 XT, and Radeon RX 7800 XT: price, VRAM, bandwidth, compute and local LLM speed compared.
| Spec | Radeon RX 9060 XT 16GB |  GeForce RTX 5060 Ti 16GB | Radeon RX 9070 XT | Radeon RX 7800 XT |
|---|---|---|---|---|
| Memory | 16 GB | 16 GBSame | 16 GBSame | 16 GBSame |
| Bandwidth | 320 GB/s | 448 GB/s40% higher | 640 GB/s100% higher | 624 GB/s95% higher |
| BF16 / FP16 compute | 103 TFLOPS | 95 TFLOPS8% lower | 195 TFLOPS89% higher | 75 TFLOPS28% lower |
| INT8 compute | 205 TOPS | 190 TOPS7% lower | 389 TOPS90% higher | 75 TOPS64% lower |
| Qwen3.8 27B speed (2-bit) | 23.6 t/s | 32.6 t/s38% faster | 45.6 t/s93% faster | 44.6 t/s89% faster |
| Largest dense model (4-bit) | 23B | 23BSame | 23BSame | 23BSame |
| Power | 160 W | 180 W13% higher | 304 W90% higher | 263 W64% higher |
| Interface | PCIe 5.0 x16 | PCIe 5.0 x8 | PCIe 5.0 x16 | PCIe 4.0 x16 |
| Released | 2025-06 | 2025-04 | 2025-03 | 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