NVIDIA RTX A6000 Pricing (June 2026)
Nathanael Chiang
The NVIDIA RTX A6000 is the Ampere-generation professional card that still earns its place in 2026: 48 GB of GDDR6 at 768 GB/s, ECC memory, and NVLink support. It is a budget route to 48 GB of VRAM for fine-tuning mid-size models and running quantized 70B-class inference, without the price of a current-gen card.
The breakdown below compares current RTX A6000 on-demand pricing.
Takeaways
- A6000 on-demand ranges from about $0.20/hr at the marketplace floor to $2.12/hr, averaging around $0.70/hr.
- Pricing has been flat year-over-year, a stable market for this card.
- It is one of the cheapest ways to get 48 GB and NVLink in a single GPU.
- Beam starts at $0.51/hr, well below the market average.
Current on-demand RTX A6000 prices (per GPU)
| Provider | SKU / Instance | On-Demand $/GPU-hr* | Notes |
|---|---|---|---|
| getdeploying floor | RTX A6000 | $0.20 | Marketplace / spot floor |
| Beam | A6000 PCIE | $0.51 | Serverless runtime, per-second billing |
| Runcrate | RTX A6000 | $0.70 | Marketplace listing |
| Market average | — | $0.70 | Across 15 tracked providers |
| Premium listings | RTX A6000 | up to $2.12 | High-availability premium |
Normalized to a single RTX A6000, even when only multi-GPU nodes are offered.
Last update: June 18, 2026
Methodology
Why you can trust these numbers:
- On-demand only: no reservations or prepaid contracts.
- Same silicon: every figure refers to the NVIDIA RTX A6000 (48 GB) SKU.
- Public price lists: pulled in June 2026 from each provider's published pricing pages.
- U.S. regions, USD: prices elsewhere vary by 5 to 20%.
A6000 vs L40S: the generational trade-off
The A6000 and L40S both give you 48 GB, but they are a generation apart. The L40S (Ada) adds FP8 support and higher bandwidth, which matters for inference throughput. The A6000 (Ampere) is older and slower, but cheaper, and it keeps NVLink, which the L40S drops. For cost-sensitive fine-tuning, experimentation, and workloads where VRAM and multi-GPU bridging matter more than FP8 speed, the A6000 is the budget pick. For pure production inference, the L40S's FP8 usually earns its small premium.
Where Beam fits
Beam's A6000 PCIE starts at $0.51/hr, below the 15-provider average of $0.70 and a fraction of the premium end. Per-second serverless billing keeps short jobs from rounding up and means idle time between runs costs nothing, which suits the experimental and fine-tuning workloads the A6000 handles best.
Why this matters for developers
- Cheap 48 GB with NVLink: the A6000 is a budget path to a full 48 GB and multi-GPU bridging.
- Stable pricing: flat year-over-year, so the main variable is provider choice.
- Know when to step up: if you need FP8 inference throughput, compare against the L40S first.
We refresh these numbers monthly.
