The RTX A6000 is the one 48 GB card on DeployGPU built on NVIDIA's Ampere architecture, the same generation as the A100, rather than Ada Lovelace. Every offer on this page is a VM you can launch now with SSH access, billed by the hour and settled to the second when you stop.
- Offers now
- 42
- From
- $0.48/GPU-hr
- GPUs per VM
- 1x, 2x, 4x
- Providers
- 4
RTX A6000 offers you can launch now
The table lists every RTX A6000 VM in stock, cheapest first, and refreshes every 15 minutes. The Launch link opens that exact offer on the deploy page, where the full hourly breakdown is shown before anything is charged.
| GPU | GPUs | Provider | Region | vCPU / RAM / Disk | Interconnect | Per GPU-hr | Per hour | Launch |
|---|---|---|---|---|---|---|---|---|
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 6 / 48 GB / 256 GB | - | $0.48 | $0.48 | Launch |
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 6 / 48 GB / 256 GB | - | $0.48 | $0.48 | Launch |
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 6 / 48 GB / 256 GB | - | $0.48 | $0.48 | Launch |
| RTX A6000 48 GB | 1x | Hyperstack | CA | 28 / 58 GB / 100 GB | - | $0.50 | $0.50 | Launch |
| RTX A6000 48 GB | 1x | Massed Compute | Iowa, US | 6 / 32 GB / 256 GB | - | $0.55 | $0.55 | Launch |
| RTX A6000 48 GB | 1x | Massed Compute | Iowa, US | 6 / 32 GB / 256 GB | - | $0.55 | $0.55 | Launch |
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 6 / 24 GB / 256 GB | - | $0.55 | $0.55 | Launch |
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 6 / 24 GB / 256 GB | - | $0.55 | $0.55 | Launch |
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 12 / 64 GB / 256 GB | - | $0.57 | $0.57 | Launch |
| RTX A6000 48 GB | 1x | Massed Compute | Iowa, US | 6 / 96 GB / 300 GB | - | $0.57 | $0.57 | Launch |
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 12 / 64 GB / 256 GB | - | $0.57 | $0.57 | Launch |
| RTX A6000 48 GB | 1x | QuantaCloud | Midwest, US | 12 / 64 GB / 256 GB | - | $0.57 | $0.57 | Launch |
For RTX A6000 prices across the whole market, compare all providers on GpuPerHour. For a reference price over time, see the RTX A6000 rental index (GPUIndexes).
What is different about this card
| Spec | RTX A6000 |
|---|---|
| Memory | 48 GB GDDR6 |
| Memory bandwidth | 768 GB/s |
| GPU-to-GPU link | NVLink bridge, 112.5 GB/s bidirectional, joins 2 GPUs |
| Low precision | No FP8 Tensor Core, the same Ampere-generation gap as the A100 |
| Power | 300 W total board power |
Specifications from NVIDIA's RTX A6000 product page. The RTX A6000 has the lowest memory bandwidth of the 48 GB cards on DeployGPU, and it is the only one of the four with an NVLink option at all, a bridge NVIDIA sells separately that links exactly 2 GPUs. Check the Interconnect column on a specific offer to see whether that link is present, since PCIe is the fallback whenever it is not.
No FP8, and 48 GB of memory
The RTX A6000 has no FP8 Tensor Cores, because it is built on the same Ampere architecture as the A100. BF16 and FP16 workloads run on it exactly as the memory numbers below suggest. Checkpoints published in FP8 or MXFP4, such as gpt-oss, still fit in memory, but treat them as fitting rather than running at native speed, the same caveat that applies to the A100 at 80 GB.
What fits in 48 GB
These are weights-only sizes of the published checkpoints, calculated rather than measured. The space left over is what vLLM has for the KV cache after its default 92% memory limit.
| Model | Format | Weights | On one RTX A6000 48 GB |
|---|---|---|---|
| gpt-oss-20b | MXFP4 | 13.8 GB | Fits, about 30 GB left, subject to the FP8/FP4 caveat above |
| Qwen3.8-27B | BF16 | 55.6 GB | Does not fit on one card. Use 2x, about 33 GB left |
| gpt-oss-120b | MXFP4 | 65.2 GB | Needs 2x, about 23 GB left |
| Gemma 4 31B | BF16 | 62.5 GB | Needs 2x, about 26 GB left |
| Llama 3.3 70B | BF16 | 141.1 GB | Needs 4x, about 36 GB left |
For a straightforward BF16 model under about 20 GB of weights, one RTX A6000 is enough. The 48 GB comparison page lines this card up against the L40, L40S and RTX 6000 Ada if bandwidth or FP8 support matters more than price for a given job.
How long until it is ready
A VM usually reaches Running in 1 to 3 minutes. App templates take about 6 to 8 minutes, because the platform waits until the app itself answers before it marks the deployment ready. On some providers the machine reports Running a little before SSH accepts connections, so a Connection refused in the first minute usually clears on its own. The Connection refused fix covers what to check if it does not.
Paid time starts when provisioning starts, but only a launch that reaches Running is charged. A launch that fails refunds the reserved hour automatically.
One GPU or eight
RTX A6000 VMs come in fixed sizes of 1, 2, 4 or 8 GPUs, and the GPUs column shows what is in stock right now. The Interconnect column says whether a specific multi-GPU VM has the NVLink bridge fitted or falls back to PCIe. One VM is the largest unit DeployGPU launches: there is no self-serve multi-node cluster. For more than 8 RTX A6000s, or a fixed term at a fixed price, request a cluster quote and a person replies by email. The multi-GPU page goes deeper on how GPUs inside one node talk to each other.
Software on the machine
In the dashboard, every RTX A6000 VM boots an Ubuntu image with NVIDIA drivers and Docker installed by the provider, and you connect over SSH as the user shown on the deployment. Three app templates set up a ready app on a private HTTPS link instead: JupyterLab with PyTorch, ComfyUI and Open WebUI with Ollama. For an OpenAI-compatible server on your own RTX A6000, follow the vLLM guide.
What you pay
The price on each row is the hourly rate you are billed, including any platform fee. Launching reserves the first hour from your credit balance, and when you stop, the unused seconds of that hour go back to the balance. Credits are bought by card with a 5% + $0.30 processing fee on each top-up. The pricing page has the full breakdown.
Stopping an RTX A6000 deletes its disk, so copy results off first. The guide to stopping a GPU without losing work shows how.
Frequently asked questions
How much does an RTX A6000 cost per hour on DeployGPU?
The table on this page lists the current hourly price of every RTX A6000 VM you can launch, cheapest first, refreshed every 15 minutes. The price shown is the rate you are billed, including any platform fee.
Does the RTX A6000 support FP8?
No. It is built on the Ampere architecture, the same generation as the A100, and NVIDIA's datasheet lists no FP8 Tensor Core throughput for it.
Does the RTX A6000 have NVLink?
It can. NVIDIA sells a bridge that links 2 RTX A6000 GPUs at 112.5 GB/s. Check the Interconnect column on a specific multi-GPU offer to see whether it is fitted.
Can I rent a single RTX A6000?
Yes, whenever a 1x offer is in stock. The GPUs column shows how many RTX A6000s come with each VM, and you pay only for the time the VM runs.
How long until I can SSH into an RTX A6000?
Most VMs reach Running in 1 to 3 minutes. On some providers SSH starts accepting connections a little after the machine reports Running, so retry for a minute before assuming something is wrong.
Is the RTX A6000 good for a first GPU deployment?
Yes. 48 GB runs most small and mid-size models, and its hourly rate, shown in the table above, makes it a low-risk place to test a template or an SSH workflow before moving to a larger GPU.
Do I need a contract to rent an RTX A6000?
No. Every RTX A6000 here is on demand with no minimum term. You prepay credits, launch, and stop whenever you want.