Rent NVIDIA B200 in the Cloud

On-demand cloud NVIDIA B200 rental rate begins at $1.99 per hour (Vultr), with 2 providers competing on rate, region, and support quality. Spot instances on RunPod drop the rate to $5.98 per hour — around up to 60% lower than on-demand — and reserved capacity can cut an additional up to 40%.

Rather than optimising for hourly rate alone, weigh spot interruption rates, egress fees, and per-second billing, which often matter more for real-world NVIDIA B200 bills.

Deploy NVIDIA B200 on Vultr (from $1.99/hr) or RunPod — check live availability and spin up in minutes.

NVIDIA B200 specifications: 192 GB HBM3e memory, 8,000 GB/s bandwidth, 2,250 FP16 TFLOPS, 75 FP32 TFLOPS, 1,000W TDP, built on Blackwell (released 2024).

For engineers evaluating the card: VRAM capacity and bandwidth are usually the two figures that determine feasibility, while TFLOPS determine time-to-result. Dense FP16 training throughput scales near-linearly with TFLOPS on well-optimised kernels, and mixed-precision pre-training can effectively double effective compute when tensor cores are in use.

Two tracked cloud providers currently offer NVIDIA B200: Vultr and RunPod. Vultr has the cheaper rate at $1.99/hr.

NVIDIA B200 delivers 2,250 FP16 TFLOPS and 75 FP32 TFLOPS, backed by 8,000 GB/s of memory bandwidth and 192 GB of VRAM. In mixed-precision fine-tuning, those numbers typically convert to solid throughput on dense models up to several tens of billions of parameters.

For low-latency inference, real-world tokens-per-second on common large language models depends more on memory bandwidth than peak FLOPS — the 8,000 GB/s figure is the relevant ceiling for autoregressive decoding. On batched workloads like diffusion image generation, compute becomes the dominant factor again.

At $1.99 per hour on the budget-friendly cloud provider, performance-per-dollar is competitive for AI-heavy workloads.

Two tracked cloud providers currently offer NVIDIA B200: Vultr and RunPod. Vultr has the cheaper rate at $1.99/hr.

To run NVIDIA B200, you have 2 cloud providers to choose from: RunPod, Vultr. The lowest-cost currently is Vultr at $1.99 per hour on-demand, with RunPod leading on spot at $5.98 per hour.

Our recommendation workflow: 1) eliminate providers without the region, compliance, or interconnect you require; 2) compare on-demand prices among the survivors; 3) add spot/reserved options for matching workloads; 4) check egress and minimum-commitment policies before signing up.

Launch a NVIDIA B200 instance on Vultr at $1.99/hr, or try RunPod for alternative regions and availability.

NVIDIA B200 is a popular choice for Next-gen AI training, large model inference, HPC. With 192 GB of memory and the efficiencies of the Blackwell generation, it delivers a capability band that covers most research and production AI work outside of the largest frontier models.

Alternatives to consider: cheaper cards for small workloads (L4, T4, RTX 4090), larger cards for frontier-scale pre-training (H100, H200, B200), and specialist cards for extreme memory needs (MI300X class). Cloud provision from $1.99 per hour on Vultr.

Vultr offers the cheapest NVIDIA B200 rate at $1.99/hr; RunPod is also available for comparison.