NVIDIA B300 SXM pricing.
prices as of · 2 providers · 9 configurations
The Blackwell Ultra refresh: the same node shape as a B200 with half again as much memory.
The cheapest B300 today is B300 · Community Cloud at RunPod, $6.94/GPU-hr, read 17 Sept 2026; the cheapest interruptible rate is $3.83/GPU-hr at Verda (DataCrunch).
Cheapest on demand
$6.94
per GPU-hour at RunPod
Median of the spread
$7.81
across 8 configurations
Cheapest interruptible
$3.83
6 of 8 publish one
Per GB of memory
$0.0241/GB-hr
market median $0.0300/GB-hr
What the board is
published by NVIDIA, not by a provider- Memory
- 288 GB
- Memory bandwidth
- 8.00 TB/s
- Architecture
- Blackwell Ultra
- Launched
- 2025
These are the chip's own figures, kept in this site's registry and the same whoever rents it out. A provider's feed is trusted for its prices, its GPU counts and the memory it states per card, never for what the board is. There is no speed score on this page: bandwidth is a published number, and turning it into a benchmark would be inventing one.
Every provider renting the B300
Cheapest per GPU-hour first. The rate is for one GPU for one hour, so a node of eight sits beside a single card at a comparable number, with the whole machine's hourly price under it. Where a provider publishes an interruptible rate it is in the same cell.
RunPodB300 · Community Cloud
1x B300 · 288 GB each · host size not stated
disk none, billed separately · regions not stated · billed per second · marketplace
Community Cloud: vetted third-party hosts, cheaper and less consistent · storage billed separately · billed per second · priced before RunPod's spec table lists it, so the memory is the board's published spec
$6.94per GPU-hour$6.94/hr all inno spot rate
Verda (DataCrunch)4B300.120V.CC · 4x B300 CC SXM6 268GB
4x B300 · 268 GB each · 120 vCPU · 1020 GB RAM · NVLink
disk none, billed separately · regions not stated · billed per hour
Dedicated Hardware Instance · storage billed separately · 1.8 TB/s between GPUs
$7.65per GPU-hour$30.60/hr all in$3.83 spot
- Verda (DataCrunch)
8B300.240V.CC · 8x B300 CC SXM6 268GB
8x B300 · 268 GB each · 240 vCPU · 2040 GB RAM · NVLink
disk none, billed separately · regions not stated · billed per hour
Dedicated Hardware Instance · storage billed separately · 1.8 TB/s between GPUs
$7.65per GPU-hour$61.20/hr all in$3.83 spot
- Verda (DataCrunch)
1B300.30V · 1x B300 SXM6 268GB
1x B300 · 268 GB each · 30 vCPU · 255 GB RAM
disk none, billed separately · regions not stated · billed per hour
Dedicated Hardware Instance · storage billed separately · 1.8 TB/s between GPUs
$7.81per GPU-hour$7.81/hr all in$3.90 spot
- Verda (DataCrunch)
2B300.60V · 2x B300 SXM6 268GB
2x B300 · 268 GB each · 60 vCPU · 510 GB RAM · NVLink
disk none, billed separately · regions not stated · billed per hour
Dedicated Hardware Instance · storage billed separately · 1.8 TB/s between GPUs
$7.81per GPU-hour$15.61/hr all in$3.90 spot
- Verda (DataCrunch)
4B300.120V · 4x B300 SXM6 268GB
4x B300 · 268 GB each · 120 vCPU · 1020 GB RAM · NVLink
disk none, billed separately · regions not stated · billed per hour
Dedicated Hardware Instance · storage billed separately · 1.8 TB/s between GPUs
$7.81per GPU-hour$31.22/hr all in$3.90 spot
- Verda (DataCrunch)
8B300.240V · 8x B300 SXM6 268GB
8x B300 · 268 GB each · 240 vCPU · 2040 GB RAM · NVLink
disk none, billed separately · regions not stated · billed per hour
Dedicated Hardware Instance · storage billed separately · 1.8 TB/s between GPUs
$7.81per GPU-hour$62.44/hr all in$3.90 spot
- RunPod
B300 · Secure Cloud
1x B300 · 288 GB each · host size not stated
disk none, billed separately · regions not stated · billed per second
Secure Cloud: RunPod's own datacentres · storage billed separately · billed per second · priced before RunPod's spec table lists it, so the memory is the board's published spec
$7.89per GPU-hour$7.89/hr all inno spot rate
Billed as a function
1 rate · never rankedThese rates buy a container that runs when it is called, not a machine you hold. CPU and memory are billed on top, a chosen region or a guarantee against preemption costs more, and the figure is a floor rather than a price. They are listed for what they are and ranked nowhere on this site.
ModalNvidia B300
1x B300 · 288 GB each · host size not stated
disk none, billed separately · regions not stated · billed per second · serverless
$0.001972 a GPU-second, the preemptible region-agnostic rate · a chosen region costs 1.15x to 1.75x and non-preemptible 3x · CPU, memory and volumes billed on top · memory is the board's published spec
$7.10per GPU-hour$7.10/hr all inno spot rate
Frequently asked
- Who rents the B300 cheapest?
- RunPod, at $6.94/GPU-hr on B300 · Community Cloud, read on 17 Sept 2026. That is the on-demand rate for one GPU for one hour; the whole machine bills $6.94/hr.
- What is the NVIDIA B300 SXM?
- The Blackwell Ultra refresh: the same node shape as a B200 with half again as much memory. It is a Blackwell Ultra board from 2025 with 288 GB of memory at 8.00 TB/s, in the SXM package. Those figures come from the vendor, not from any provider's price page.
- Can I rent a B300 on spot?
- Yes. 6 of the 8 configurations here publish an interruptible rate, the cheapest being $3.83/GPU-hr at Verda (DataCrunch), 50% under its own on-demand rate. The machine can be taken back mid-run, so it suits work that checkpoints.
- Why do B300 prices differ so much?
- The spread here runs from $6.94/GPU-hr to $7.89/GPU-hr for the same silicon, because what else comes with it differs: the vCPU and RAM around the GPU, whether a disk is included, how many regions it runs in, and whether the machine belongs to the provider or to a third party on a marketplace. Each row says which.
- How current are these prices?
- They are read every morning from each provider's own public page or API, and this page recomputes with them. Last refresh: 17 Sept 2026. Reserved and committed rates are not indexed. Prices exclude tax.