RTX A6000 Price & Used Prices: Which Local LLMs Can It Run?
Price alert
We'll email you when the Xianyu (CN) price reaches your target.
The RTX A6000 has 48GB of GDDR6 ECC memory with 768 GB/s of bandwidth. The median asking price is about $4,612 on Xianyu. At 4-bit quantization and 8K context it fits a dense model of up to about 79B parameters, at roughly 12 tokens/s.
Xianyu
RTX A6000 price history
Xianyu
Buy links may carry affiliate tags; they do not change the price you pay.
Full RTX A6000 specifications
- Memory48 GB · GDDR6 ECC
- Bandwidth768 GB/s
- FP16 / BF16154.9 TFLOPS
- INT8309.7 TOPS
- Bus width
- 384-bit
- Architecture
- Ampere (GA102)
- Cores
- 10,752
- Tensor cores
- 336
- Ecosystem
- CUDA
- Interface
- PCIe 4.0 x16
- NVLink
- 112.5 GB/s
- Power
- 300 W
- Power connector
- 8-pin EPS
- Slots
- 2
- Length
- 267 mm
- Type
- Workstation
- Launch date
- Oct 5, 2020
What LLMs can the RTX A6000 run?
Models up to 50B parameters on a single RTX A6000, versions released since 2026 only
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
 Gemma 4 E4B8B | 3.8 GB | Runs | 130.4 | 128K |
 Qwen3.5 9B9.7B | 4.1 GB | Runs | 120.5 | 256K |
| Gemma 4 12B12B | 4.7 GB | Runs | 104.9 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 10.5 GB | Runs | 140.2 | 256K |
| Qwen3.6 27B27.8B | 11.8 GB | Runs | 45.1 | 256K |
| Qwen3.8 27B27.8B | 9.8 GB | Runs | 53.7 | 256K |
 GLM 4.7 Flash30B (3B active) | 11.9 GB | Runs | 154 | 198K |
| Gemma 4 31B30.7B | 11.8 GB | Runs | 43.5 | 256K |
| Qwen3.6 35B A3B36B (3B active) | 12.3 GB | Runs | 167.7 | 256K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 5 GB | Runs | 102.1 | 128K |
| Qwen3.5 9B9.7B | 5.7 GB | Runs | 89.8 | 256K |
| Gemma 4 12B12B | 7.1 GB | Runs | 72.5 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 16.9 GB | Runs | 113.1 | 256K |
| Qwen3.6 27B27.8B | 16.8 GB | Runs | 32.2 | 256K |
| Qwen3.8 27B27.8B | 16.5 GB | Runs | 32.8 | 256K |
| GLM 4.7 Flash30B (3B active) | 18.3 GB | Runs | 131.1 | 198K |
| Gemma 4 31B30.7B | 18.3 GB | Runs | 29 | 256K |
| Qwen3.6 35B A3B36B (3B active) | 22.1 GB | Runs | 134.9 | 256K |
| Model | Size | Runs? | Speed t/s | Max context |
|---|---|---|---|---|
| Gemma 4 E4B8B | 8.2 GB | Runs | 64.7 | 128K |
| Qwen3.5 9B9.7B | 9.5 GB | Runs | 56 | 256K |
| Gemma 4 12B12B | 12.7 GB | Runs | 42.2 | 256K |
| Gemma 4 26B A4B25.2B (3.8B active) | 26.9 GB | Runs | 86.9 | 256K |
| Qwen3.6 27B27.8B | 28.6 GB | Runs | 19.3 | 256K |
| Qwen3.8 27B27.8B | 29 GB | Runs | 19 | 256K |
| GLM 4.7 Flash30B (3B active) | 31.8 GB | Runs | 99.8 | 198K |
| Gemma 4 31B30.7B | 32.6 GB | Runs | 16.7 | 202K |
| Qwen3.6 35B A3B36B (3B active) | 36.9 GB | Runs | 104.2 | 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.
Dual, 4x and 8x RTX A6000 for LLMs: which models fit?
Cards needed for 50B+ models released since 2026, with vLLM tensor parallelism at 32K context
| Model | Size | Cards | Total price | Total VRAM | Speed t/s | Total power |
|---|---|---|---|---|---|---|
| Qwen3.8 Flash Next176B (6B active) | 78.9 GB | 2 cards | $9,224 | 96 GB | 105.7 | 600 W |
 DeepSeek V4 Flash284B (13B active) | 96.8 GB | 4 cards | $18,448 | 192 GB | 83.3 | 1,200 W |
 Hy3295B (21B active) | 107.2 GB | 4 cards | $18,448 | 192 GB | 37.3 | 1,200 W |
| GLM 5.3 Flash320B (18B active) | 108.7 GB | 4 cards | $18,448 | 192 GB | 66 | 1,200 W |
| GLM 5.3744B (40B active) | 253.9 GB | 8 cards | $36,896 | 384 GB | 32.5 | 2,400 W |
| Hy4 preview770B (49B active) | 286.1 GBest. | 8 cards | $36,896 | 384 GB | 24.3 | 2,400 W |
 Kimi K2.7 Code1,000B (32B active) | 339.5 GB | 8 cards | $36,896 | 384 GB | 39.6 | 2,400 W |
| Model | Size | Cards | Total price | Total VRAM | Speed t/s | Total power |
|---|---|---|---|---|---|---|
| Qwen3.8 Flash Next176B (6B active) | 111.3 GB | 4 cards | $18,448 | 192 GB | 87.6 | 1,200 W |
| DeepSeek V4 Flash284B (13B active) | 155.1 GB | 4 cards | $18,448 | 192 GB | 59.8 | 1,200 W |
| Hy3295B (21B active) | 182.2 GB | 8 cards | $36,896 | 384 GB | 27.6 | 2,400 W |
| GLM 5.3 Flash320B (18B active) | 199.7 GB | 8 cards | $36,896 | 384 GB | 41.3 | 2,400 W |
| Model | Size | Cards | Total price | Total VRAM | Speed t/s | Total power |
|---|---|---|---|---|---|---|
| DeepSeek V4 Flash284B (13B active) | 161.9 GB | 4 cards | $18,448 | 192 GB | 57.9 | 1,200 W |
| Qwen3.8 Flash Next176B (6B active) | 188.2 GB | 8 cards | $36,896 | 384 GB | 62.3 | 2,400 W |
| Hy3295B (21B active) | 317.7 GB | 8 cards | $36,896 | 384 GB | 18.7 | 2,400 W |
| GLM 5.3 Flash320B (18B active) | 341 GB | 8 cards | $36,896 | 384 GB | 26.1 | 2,400 W |
Card counts are 1 / 2 / 4 / 8, each using 90% of its VRAM. Speed is based on one card’s bandwidth; price covers GPUs only. Supports NVLink (112.5 GB/s). Whether it runs also depends on vLLM support.
RTX A6000 rental price per hour
On-demand rental of the RTX A6000 starts at $0.33 per hour (RunPod), about $80/month at 8 hours a day. Buying one (Xianyu median $4,612) pays off in about 11.4 years, 5.7 years or 1.9 years at 4, 8 or 24 hours a day.
| Option | Price | 4 h/day | 8 h/day | 24 h/day | View |
|---|---|---|---|---|---|
 RunPodRentUpdated 22 min. ago | $0.33/hr | $40/mo | $80/mo | $241/mo | Rent |
 Vast.aiRentUpdated 22 min. ago | $0.43/hr | $52/mo | $104/mo | $313/mo | Rent |
 XianyuUpdated 21 min. ago | $4,612 | Payback ~11.4 yrsPower $7/mo | Payback ~5.7 yrsPower $13/mo | Payback ~1.9 yrsPower $39/mo | Buy |
Break-even uses the US residential average ($0.18/kWh) and 300 W rated power, excluding resale value and the rest of the build.
Other markets
eBay
Amazon.jpBuy links may carry affiliate tags; they do not change the price you pay.
FAQ
How large a model can the RTX A6000 run?
A single RTX A6000 has 48GB of GDDR6 ECC VRAM. At 4-bit quantization and 8K context it fits a dense model of up to about 79B parameters, or about 43B at 8-bit.
Can the RTX A6000 run a 70B model?
Yes. At 4-bit with an 8K context a 70B dense model needs about 43GB, which fits in the RTX A6000's 48GB.
What are the best LLMs to run on the RTX A6000?
Among models released in 2026, the largest dense model that fits entirely in VRAM at 4-bit is Gemma 4 31B, at an estimated 29 tokens/s. Among MoE models, Qwen3.6 35B A3B is the fastest at about 134.9 tokens/s. The full list is in the table above.
How many tokens per second does the RTX A6000 get on LLMs?
Estimated from memory bandwidth at 4-bit and 8K context: 8B dense: about 97.9 tokens/s, 14B dense: about 59 tokens/s, 32B dense: about 27.3 tokens/s, and 70B dense: about 12.9 tokens/s. MoE models read only the active parameters for each token, so they run much faster. Real-world results are usually 70%–100% of the estimate.
What LLMs can dual RTX A6000 cards run?
With vLLM tensor parallelism at 4-bit and 32K context, two RTX A6000 cards cannot fit any model of 50B or larger released in 2026. At 4-bit with 32K context the minimum is 4 cards, which fits Qwen3.8 Flash Next and DeepSeek V4 Flash.
How much does a RTX A6000 cost?
As of Sep 29, 2026: eBay median $5,195 (19 listings), Amazon.jp median $8,877 (5 listings), and Xianyu median $4,612 (16 listings). Figures are medians of live listings, refreshed every 6 hours.
Is the RTX A6000 price going up or down?
As of Sep 29, 2026, the Xianyu median is down 3.1% over 7 days, down 4.6% over 30 days.
How much does it cost to rent a RTX A6000 per hour?
As of Sep 29, 2026: RunPod at $0.33 per hour and Vast.ai at $0.43 per hour. All are single-GPU on-demand rates, excluding storage and data transfer.
Is it cheaper to buy or rent a RTX A6000 at 8 hours a day?
Taking the Xianyu median of $4,612 and electricity at $0.18 per kWh, at 8 hours a day buying pays for itself in about 5.7 years (versus RunPod at $0.33 per hour). If you will use it for longer than 5.7 years, buy; otherwise, rent.
RTX A6000 vs Radeon PRO W7900: which is better for LLMs?
The Radeon PRO W7900 has 48GB of VRAM and 864 GB/s of bandwidth; the RTX A6000 has 48GB and 768 GB/s. At the Xianyu median, the Radeon PRO W7900 costs about $3,050, 33.9% less than the RTX A6000 ($4,612). For 70B dense models at 4-bit, the RTX A6000 is estimated at about 12.9 tokens/s and the Radeon PRO W7900 at about 10.4 tokens/s.