Loading prices…
B300 SXM rental prices
NVIDIA · Blackwell Ultra · 288 GB HBM3e · SXM · 2025
17 providers · 45 configurations · prices as of Nov 10, 2026, 23:02 UTC
Lowest on-demand
$5.60per GPU·h · ≈ $4,085/month
at SeewebMedian on-demand
$9.25/GPU·h
across 16 providers
Lowest spot
$2.51/GPU·h
Reserved from $4.75
FP16/BF16 cost
$2.49/PFLOP·h
#35 of 75 GPUs for value
Providers
17
45 configurations
Price history
Daily B300 SXM prices per GPU-hour. The shaded period is reconstructed from public archives: the version history of SkyPilot's price catalogs, Internet Archive copies of providers' pricing pages and the open gpu-price-feed project (fewer providers than live data). After that, PetaPrice's own tracking.
Lowest on-demandMedian on-demandLowest spot
Price by provider
Cheapest On-demand B300 SXM offer at each provider, per GPU-hour. Shorter and greener is cheaper.
Seeweb🇮🇹$5.60best
Massed Compute🇺🇸$6.60+18%
RunPod🇺🇸$6.94+24%
Hyperstack🇬🇧$7.40+32%
Civo🇬🇧$7.49+34%
UpCloud🇫🇮$7.99+43%
Arkane Cloud🇫🇷$8.99+61%
Scaleway🇫🇷$9.01+61%
Nebius🇳🇱$9.50+70%
DigitalOcean🇺🇸$11.2+100%
Lium$11.92.1×
Verda (DataCrunch)🇫🇮$11.92.1×
Oracle Cloud🇺🇸$15.02.7×
Latitude.sh🇺🇸$16.02.9×
AWS🇺🇸$17.83.2×
EmpirioLabs AI🇺🇸$23.84.3×
Thin vertical line = median across providers ($9.01).
Every provider we track
Pricing
Size
Host
Provider HQ
17 providers · 45 configurations
Sort
- SeewebOn-demandReserved🇮🇹 HQ in Italyvia Seeweb pricing pagefrom · On-demandCheapest$5.60/GPU·hCheapest
- Massed ComputeOn-demand🇺🇸 HQ in United Statesvia Massed Compute pricing pagefrom · On-demand+18%$6.60/GPU·h+18%
- 8× GPU280 vCPU · 2,800 GB RAM · 6,000 GB disk$6.60 /GPU·h$52.8 /hOn-demandRent ↗
- RunPodcommunityOn-demand🇺🇸 HQ in United Statesvia RunPodfrom · On-demand+24%$6.94/GPU·h+24%
- HyperstackOn-demand🇬🇧 HQ in United Kingdomvia Hyperstack pricing pagefrom · On-demand+32%$7.40/GPU·h+32%
- 8× GPU—$7.40 /GPU·h$59.2 /hOn-demandRent ↗
- CivoOn-demandReserved🇬🇧 HQ in United Kingdomvia Civo pricing pagefrom · On-demand+34%$7.49/GPU·h+34%
- UpCloudOn-demandSpot🇫🇮 HQ in Finlandvia UpCloud pricing pagefrom · On-demand+43%$7.99/GPU·h+43%
- Arkane CloudOn-demand🇫🇷 HQ in Francevia Arkane Cloud pricing pagefrom · On-demand+61%$8.99/GPU·h+61%
- 1× GPU30 vCPU · 275 GB RAM$8.99 /GPU·h$8.99 /hOn-demandParis / LyonRent ↗
- ScalewayOn-demand🇫🇷 HQ in Francevia Shadeformfrom · On-demand+61%$9.01/GPU·h+61%
- 8× GPU · NVLink224 vCPU · 3,750 GB RAM · 6,000 GB disk$9.01 /GPU·h$72.1 /hOn-demand🇫🇷 FR, Paris○ sold outRent ↗
What can the B300 SXM run?
Memory for the weights plus one 32K-token sequence of context, how many GPUs hold it, and what that replica costs at the cheapest in-stock on-demand rate.
| Model | Params | Memory | GPUs | $/hr | $/mo |
|---|---|---|---|---|---|
gpt-oss-20b OpenAI · 20.9B | 20.9B · 3.6B active | 20 GiB | 1 | $5.60 at Seeweb | $4,085 |
Gemma 3 27B Google · 27.4B | 27.4B | 28 GiB | 1 | $5.60 at Seeweb | $4,085 |
Qwen3-32B Alibaba · 32.8B | 32.8B | 39 GiB | 1 | $5.60 at Seeweb | $4,085 |
Llama 3.3 70B Instruct Meta · 70.6B | 70.6B | 76 GiB | 1 | $5.60 at Seeweb | $4,085 |
gpt-oss-120b OpenAI · 117B | 117B · 5.1B active | 110 GiB | 1 | $5.60 at Seeweb | $4,085 |
Qwen3-235B-A22B Alibaba · 235B | 235B · 22B active | 225 GiB | 1 | $5.60 at Seeweb | $4,085 |
DeepSeek-R1 DeepSeek · 684B | 684B · 37B active | 640 GiB | 4 | $22.38 at Seeweb | $16,340 |
Kimi K3 Moonshot AI · 2.78T | 2.78T · 104B active | 2.53 TiB | 16 | $89.53 at Seeweb | $65,359 |
Assumes 90% of each GPU's memory is usable and rounds up to 1, 2, 4, 8… GPUs for tensor parallelism (up to 64). $/hr is the GPU count times the cheapest in-stock on-demand rate per GPU; check that the provider rents instances that large. Serving many requests at once needs more memory for context.
Alternatives to the B300 SXM
GPUs with similar compute and memory, and their lowest on-demand price right now.
Instinct MI350X288 GB · 2.31 PF FP16$6.16/h$2.67/PFLOP·h- GB300288 GB · 2.5 PF FP16$13.7/h$5.49/PFLOP·h
- Instinct MI355X288 GB · 2.52 PF FP16$8.60/h$3.42/PFLOP·h
- B200 SXM180 GB · 2.25 PF FP16$5.43/h$2.41/PFLOP·h
- GB200186 GB · 2.5 PF FP16$10.5/h$4.20/PFLOP·h
- Instinct MI325X256 GB · 1.31 PF FP16$3.80/h$2.91/PFLOP·h
B300 SXM pricing questions
How much does it cost to rent an B300 SXM?
On-demand B300 SXM prices start at $5.60 per GPU-hour (Seeweb). The median across 16 providers is $9.25, and the most expensive is $23.8. Running one GPU around the clock costs about $4,085 per month at the lowest price.
Is there cheaper spot or interruptible B300 SXM capacity?
Yes. Spot (interruptible) B300 SXM capacity starts at $2.51 per GPU-hour, 55% below the cheapest on-demand price. Spot instances can be preempted at any time, so use them for checkpointed training or batch jobs.
How much compute do you get per dollar with the B300 SXM?
At $5.60 per hour and 2.25 PF of dense FP16/BF16, the B300 SXM costs $2.49 per PFLOP-hour. That ranks 35 of 75 GPUs with on-demand prices on PetaPrice. Figures are datasheet peaks; real utilization is typically 30–60%.
What are the B300 SXM's key specs?
288 GB HBM3e with 8,000 GB/s of memory bandwidth, Blackwell Ultra architecture (2025), SXM form factor, 1100 W. Dense peak: 2.25 PF FP16 / BF16, 4.5 PF FP8, 13.5 PF FP4, 0.33 PF INT8, 1.1 PF TF32, 0.075 PF FP32, 0.0013 PF FP64.
How current are these prices?
Prices were collected on Nov 10, 2026, 23:02 UTC from provider APIs and public catalogs and refresh about every 2 hours. They are indicative list prices; always confirm on the provider's website before booking.