NVIDIA RTX PRO 6000 Pricing
Eli Mernit
Renting an NVIDIA RTX PRO 6000 costs $1.09 to $4.50 per GPU-hour on demand for a full card across the providers below, and the getdeploying median is $2.18 across 41 providers. Beam's is $1.09/hr with vCPU, RAM, and NVMe included.
Key takeaways
- Beam's $1.09/hr is the lowest full-card on-demand price in the table and half the $2.18 median.
- Stock is patchy: getdeploying shows Hyperstack's on-demand flavor and all Sesterce configs out of stock.
- Off the table, getdeploying's cheapest verified in-stock on-demand listing is $0.66/hr (Packet·ai), and Vast.ai spot starts at $0.33/hr.
RTX PRO 6000 cloud prices per hour, by provider
| Provider | SKU | On-demand $/GPU-hr | Notes |
|---|---|---|---|
| Beam | RTX PRO 6000 PCIe | $1.09 | On-demand machine; vCPU, RAM, and NVMe included |
| RunPod | RTX PRO 6000, Community Cloud | $1.69 | Secure Cloud $2.09; in stock |
| Hyperstack | RTX PRO 6000 Server Edition | $1.85 | Spot $1.48; out of stock |
| Verda | RTX PRO 6000 96GB | $1.98 | Spot $0.99; in stock |
| Sesterce | RTX PRO 6000 | $2.41 | Same listing on Runcrate; out of stock |
| AWS | g7e.2xlarge | $3.36 | 8 vCPU, 64 GiB; us-east-1 |
| Google Cloud | g4-standard-48 | $4.50 | 48 vCPU, 180 GiB; us-central1; spot $1.68 |
| Market median | — | $2.18 | Across 41 providers with an on-demand price (getdeploying) |
How these prices were collected
Prices are on-demand, per-GPU list prices in US dollars for one whole card, with spot excluded from the price column and stock status from getdeploying. Fractional shapes are left out, such as Google Cloud's g4-standard-6 (one-eighth of a GPU for $0.65/hr) and RunPod's 48 GB MIG slice ($1.09/hr). Marketplace data from Vast.ai, via getdeploying, supplies the spot floor.
- Verda's pricing page for on-demand and spot rates
- Hyperstack's GPU pricing page for on-demand and spot
- RunPod's GPU pricing page for Community and Secure Cloud
- Sesterce's compute page for its on-demand VM rate
- AWS's EC2 On-Demand pricing page for g7e instances
- Google Cloud's accelerator-optimized VM pricing for G4 shapes
- getdeploying's RTX PRO 6000 price tracker for stock and the median
Why the RTX PRO 6000 is priced the way it is
Memory is the draw. The RTX PRO 6000 Blackwell carries 96 GB of GDDR7 on a PCIe card, twice an L40S and more than an 80 GB A100, without the HBM of data-center training GPUs. The market prices it in between: its $2.18 median sits below the H100's but above the A100's $1.77 and the L40S's $1.54, which our L40S price comparison breaks down.
Prices spread widely because stock is uneven across the 54 providers getdeploying tracks and hyperscalers bundle the card into large VMs.
Where to rent an RTX PRO 6000 for less
Beam's $1.09/hr is the lowest full-card rate in the table, not the market floor. The cheapest verified in-stock on-demand listing on getdeploying is Packet·ai at $0.66/hr. Before booking a listing that cheap, confirm it's a whole 96 GB card and check what CPU and RAM come with it; Beam's machine price includes vCPU, RAM, and NVMe. Spot is lower still on Vast.ai, from $0.33/hr, and Verda's $0.99/hr spot rate also undercuts Beam; both can be interrupted.
What this means for your workload
Pick it when a model outgrows a 48 GB card but doesn't need a multi-GPU training node. At two bytes per parameter, 96 GB holds about 48B parameters of FP16 weights, and FP8 or Blackwell's FP4 support stretches that further. It suits single-GPU inference, such as an open model behind an OpenAI-compatible API, and image or video generation.
FAQ
How much does an RTX PRO 6000 cost per hour?
Full-card on-demand prices in the table run from $1.09 to $4.50 per GPU-hour. The getdeploying median across 41 providers is $2.18, and Beam's RTX PRO 6000 machine is $1.09/hr.
What's the cheapest way to rent an RTX PRO 6000?
Spot is cheapest, from $0.33/hr on Vast.ai, if your job can handle interruptions. On demand, the lowest verified in-stock listing on getdeploying is $0.66/hr from Packet·ai; check that it's a whole card.
RTX PRO 6000 vs L40S: which should I rent?
Rent the L40S if your model fits in 48 GB, since its median is $1.54 against $2.18 and Beam charges $0.76/hr against $1.09/hr. Rent the RTX PRO 6000 when you need 96 GB on one card.
