New B200 spot capacity is live in US East from $1.69 per GPU-hour. See availability
Comparison

B200 SXM vs H200 SXM

Both run on the same nodes, the same templates and the same tiers here — so the choice comes down to memory, bandwidth and price. Every figure below is from the catalogue or the vendor datasheet.

B200
$1.69/h180 GB · Blackwell · 16 available
H200
$0.75/h141 GB · Hopper · 32 available
Price gap
56%On spot, per GPU-hour
Over 100 h
$94.00Difference for the same GPU-hours
In short

What actually separates them.

Four differences that change a decision, computed from the numbers rather than asserted.

Bigger models fit on the B200

180 GB against 141 GB. At FP16 that is about 72B parameters on one GPU, against 56B.

The H200 costs 56% less per hour

$0.75 against $1.69 on spot — $94.00 of difference over 100 GPU-hours.

Token generation is faster on the B200

7,700 GB/s against 4,800 GB/s. Inference reads the whole model from memory for every token, so bandwidth, not FLOPS, usually sets the ceiling.

More memory per dollar on the H200

188 GB per dollar per hour against 107 GB. The usual tie-breaker when both cards fit the job.

Side by side

Every figure we hold on both cards.

A row is marked only when both values exist and one is genuinely better. B200 leads on 4 rows, H200 on 5.

  B200 SXM H200 SXM
GPU memoryDecides which models fit at all 180 GB HBM3e 141 GB HBM3e
Memory bandwidthSets token throughput on inference 7,700 GB/s 4,800 GB/s
FP32 75 TFLOPS 67 TFLOPS
8-bit float (FP8)Halves the memory a model needs, where the stack supports it Supported Supported
Architecture Blackwell (2025) Hopper (2024)
Board powerLower is cheaper to run at scale — for us, not for your bill 1000 W 700 W
GPU-to-GPU linkOnly matters for multi-GPU training NVLink NVLink
Node sizes 1×, 2×, 4×, 8× 1×, 2×, 4×, 8×
Spot price $1.69/h $0.75/h
On-demand price $2.49/h $1.09/h
Memory per dollarGPU memory per $1 of spot time, per hour 107 GB 188 GB
30-day reclaim rateShare of spot instances reclaimed 5–10% 5–10%
Available now 16 GPUs 32 GPUs

Specifications come from each vendor's datasheet — linked on the B200 page and the H200 page. Prices are ours, per GPU-hour, billed per minute.

What fits

The same models, on one GPU of each.

Weights only, with 20% of the memory kept free for activations and the KV cache. A cross means one GPU is not enough — the model page has the multi-GPU arithmetic.

Model Precision B200 180 GB H200 141 GB
Llama 3.1 8B8B parameters FP16 · 16 GB Fits Fits
FP8 · 8 GB Fits Fits
INT4 · 4 GB Fits Fits
Qwen2.5 32B32B parameters FP16 · 64 GB Fits Fits
FP8 · 32 GB Fits Fits
INT4 · 16 GB Fits Fits
Llama 3.3 70B70B parameters FP16 · 140 GB Fits Needs 2× node
FP8 · 70 GB Fits Fits
INT4 · 35 GB Fits Fits
Mixtral 8x22B141B parameters FP16 · 282 GB Needs 2× node Needs 4× node
FP8 · 141 GB Fits Needs 2× node
INT4 · 71 GB Fits Fits
Cost

The same 100 GPU-hours on each.

Compute only, billed per minute. Storage is $0.08 per GB-month on either card.

 B200H200
One hour, spot$1.69$0.75
One day, spot$40.56$18.00
100 GPU-hours, spot$169.00$75.00
100 GPU-hours, on-demand$249.00$109.00
A month, reserved$1,350.50$576.70

Take the B200 when…

  • Training and full fine-tuning — Enough memory for optimiser states and activations on models a smaller card can only run in inference.
  • Serving large models — A 70B model at FP8 on a single GPU, with room for a long context window.
  • Multi-GPU jobs over NVLink — GPU-to-GPU traffic stays off the PCIe bus, which is what makes tensor and pipeline parallelism worth it.

Full B200 page

Take the H200 when…

  • Training and full fine-tuning — Enough memory for optimiser states and activations on models a smaller card can only run in inference.
  • Serving large models — A 70B model at FP8 on a single GPU, with room for a long context window.
  • Multi-GPU jobs over NVLink — GPU-to-GPU traffic stays off the PCIe bus, which is what makes tensor and pipeline parallelism worth it.

Full H200 page

FAQ

B200 or H200?

Which is cheaper, the B200 or the H200?

The H200 SXM, at $0.75 per GPU-hour on spot against $1.69 — 56% less. On-demand: $1.09 against $2.49.

Which one should I take for inference?

Take the one that holds the model with room for the KV cache, then the one with more memory bandwidth. Here that is the B200 on capacity (180 GB) and the B200 on bandwidth.

Can I run both on the same job?

Not inside one instance — a node holds GPUs of a single model. You can run two instances in parallel, on the same persistent volume in turn, or split the work: a big card for the model that needs capacity, a cheap one for the parts that do not.

Other comparisons: the full list · all 47 models

Get started

Try both for less than the price of deciding.

An hour on each costs $2.44 on spot. Nothing renews, and your disk moves between them.

Billed per minute from the moment the instance is reachable. Minimum credit $40, no subscription.