Skip to content
GPU Finder

Runpod vs Vast GPU Pricing

Side-by-side GPU cloud pricing comparison with live data and real-time availability.

Full catalogs: Runpod GPU pricing · Vast GPU pricing

Runpod:36 GPUs
Vast:52 GPUs
Shared GPUs:31
Updated:

Vast is cheaper on 30 of 31 shared GPU models

No data
$1.60/GPU/hrH100 on-demand
on Vast
Visit Vast
Runpod H100: $1.99/GPU/hr · Runpod and Vast can both be cost-effective for flexible GPU workloads, but their discounted capacity works differently. Runpod exposes explicit spot pricing fields, while Vast is a marketplace with interruptible and bid-related pricing. GPU Finder compares both with live availability and 30-day reliability context so the cheapest row is not evaluated in isolation.

Runpod spot vs Vast interruptible pricing

Runpod and Vast can both be cost-effective for flexible GPU workloads, but their discounted capacity works differently. Runpod exposes explicit spot pricing fields, while Vast is a marketplace with interruptible and bid-related pricing. GPU Finder compares both with live availability and 30-day reliability context so the cheapest row is not evaluated in isolation.

ChooseWhen it fitsWatch out for
Runpod spotYou want explicit provider-sourced spot fields and a simpler Secure/Community Cloud distinction.Confirm the spot row exists for the GPU you need and compare against on-demand.
Vast interruptible/bidYou want marketplace floor pricing and can tolerate host variability.Bid floors, bandwidth/storage costs, and host reliability can change the real cost/risk.
On-demand insteadYou need predictable runtime or cannot checkpoint.Higher $/GPU-hour may be worth it when interruption risk is unacceptable.

Runpod spot floor

Check the live table for current Runpod spot rows.

Runpod spot GPU pricing

Vast interruptible/bid floor

Vast interruptible currently from $0.01/GPU-hr on RTX A4000.

Vast interruptible GPU pricing
Compare GPU models:B200H100A100H200

GPU pricing comparison

  • B200
    No dataNo data
    $5.98/GPU/hr · Runpod
    VisitVisit
    Runpod $5.98/GPU/hrVast $6.25/GPU/hrΔ $0.27 (4%)Runpod cheaper
  • RTX A5000
    No dataNo data
    $0.14/GPU/hr · Vast
    VisitVisit
    Runpod $0.16/GPU/hrVast $0.14/GPU/hrΔ $0.02 (15%)Vast cheaper
  • RTX 4000 Ada
    No dataNo data
    $0.16/GPU/hr · Vast
    VisitVisit
    Runpod $0.20/GPU/hrVast $0.16/GPU/hrΔ $0.04 (18%)Vast cheaper
  • RTX A6000
    No dataNo data
    $0.28/GPU/hr · Vast
    VisitVisit
    Runpod $0.33/GPU/hrVast $0.28/GPU/hrΔ $0.05 (15%)Vast cheaper
  • RTX A4500
    No dataNo data
    $0.11/GPU/hr · Vast
    VisitVisit
    Runpod $0.19/GPU/hrVast $0.11/GPU/hrΔ $0.08 (43%)Vast cheaper
  • RTX A2000
    No dataNo data
    $0.03/GPU/hr · Vast
    VisitVisit
    Runpod $0.12/GPU/hrVast $0.03/GPU/hrΔ $0.09 (72%)Vast cheaper
  • RTX 3070
    No dataNo data
    $0.04/GPU/hr · Vast
    VisitVisit
    Runpod $0.13/GPU/hrVast $0.04/GPU/hrΔ $0.09 (68%)Vast cheaper
  • RTX 4080
    No dataNo data
    $0.18/GPU/hr · Vast
    VisitVisit
    Runpod $0.27/GPU/hrVast $0.18/GPU/hrΔ $0.09 (33%)Vast cheaper
  • RTX 3080
    No dataNo data
    $0.08/GPU/hr · Vast
    VisitVisit
    Runpod $0.17/GPU/hrVast $0.08/GPU/hrΔ $0.09 (53%)Vast cheaper
  • RTX 3080 Ti
    No dataNo data
    $0.08/GPU/hr · Vast
    VisitVisit
    Runpod $0.18/GPU/hrVast $0.08/GPU/hrΔ $0.10 (54%)Vast cheaper
  • RTX A4000
    No dataNo data
    $0.07/GPU/hr · Vast
    VisitVisit
    Runpod $0.17/GPU/hrVast $0.07/GPU/hrΔ $0.10 (58%)Vast cheaper
  • RTX 4070 Ti
    No dataNo data
    $0.07/GPU/hr · Vast
    VisitVisit
    Runpod $0.19/GPU/hrVast $0.07/GPU/hrΔ $0.12 (64%)Vast cheaper
  • RTX 3090 Ti
    No dataNo data
    $0.14/GPU/hr · Vast
    VisitVisit
    Runpod $0.27/GPU/hrVast $0.14/GPU/hrΔ $0.13 (50%)Vast cheaper
  • RTX 4090
    No dataNo data
    $0.20/GPU/hr · Vast
    VisitVisit
    Runpod $0.34/GPU/hrVast $0.20/GPU/hrΔ $0.14 (41%)Vast cheaper
  • RTX 3090
    No dataNo data
    $0.08/GPU/hr · Vast
    VisitVisit
    Runpod $0.22/GPU/hrVast $0.08/GPU/hrΔ $0.14 (63%)Vast cheaper
  • V100
    No dataNo data
    $0.03/GPU/hr · Vast
    VisitVisit
    Runpod $0.19/GPU/hrVast $0.03/GPU/hrΔ $0.16 (86%)Vast cheaper
  • A40
    No dataNo data
    $0.28/GPU/hr · Vast
    VisitVisit
    Runpod $0.49/GPU/hrVast $0.28/GPU/hrΔ $0.21 (43%)Vast cheaper
  • RTX 5080
    No dataNo data
    $0.13/GPU/hr · Vast
    VisitVisit
    Runpod $0.39/GPU/hrVast $0.13/GPU/hrΔ $0.26 (65%)Vast cheaper
  • RTX 6000 Ada
    No dataNo data
    $0.48/GPU/hr · Vast
    VisitVisit
    Runpod $0.74/GPU/hrVast $0.48/GPU/hrΔ $0.26 (35%)Vast cheaper
  • RTX PRO 5000
    No dataNo data
    $0.55/GPU/hr · Vast
    VisitVisit
    Runpod $0.82/GPU/hrVast $0.55/GPU/hrΔ $0.27 (33%)Vast cheaper

Egress Cost Comparison

Runpod

Free

Vast

Free

Exclusive GPUs

Only on Runpod (5)

  • B300
  • MI300X
  • RTX 2000 Ada
  • RTX 4000 SFF Ada
  • RTX 5000 Ada

Only on Vast (21)

  • A10
  • A16
  • A30
  • P4
  • Quadro RTX 6000
  • Quadro RTX 8000
  • RTX 2060
  • RTX 2070
  • RTX 2080 Ti
  • RTX 3060
  • RTX 3060 Ti
  • RTX 3070 Ti
  • RTX 4060
  • RTX 4060 Ti
  • RTX 4070
  • RTX 5060
  • RTX 5060 Ti
  • RTX 5070 Ti
  • Tesla P40
  • Titan RTX
  • Titan V

Which Should You Choose?

Choose Runpod if...

  • You want explicit provider-sourced spot fields and a simpler Secure/Community Cloud distinction
  • You prefer a managed cloud experience with clearer reliability guarantees
  • You need serverless GPU endpoints or curated high-end training clusters

Choose Vast if...

  • You want marketplace floor pricing and can tolerate host variability
  • You are running batch jobs or experiments where occasional interruptions are acceptable
  • You need access to a massive variety of GPU types and configurations

Frequently Asked Questions

Is Runpod spot the same as Vast interruptible?

No. Runpod exposes explicit spot fields; Vast uses marketplace interruptible/bid semantics.

Which is cheaper, Runpod spot or Vast interruptible?

The cheapest provider can change by GPU model, stock, and bid state; compare current rows on the page.

What GPU models should I compare first?

Start with H100/H200 for high-end training, A100 for legacy value, and B200 where verified current rows exist.

How does reliability affect spot GPU pricing?

A lower price is less useful if stock is rarely available or the host is interruption-prone; compare the price with 30-day reliability.

Best for your workload

Training — Runpod is our cost pickInference — Runpod is our ops pickFine-tuning — Runpod is our cost pickBudget — Runpod is our ops pickConsumer / Local LLM — Runpod is our performance pick

Related Comparisons

AWSvsDigital Ocean
Digital OceanvsLambda
Digital OceanvsRunpod
Digital OceanvsVast
LambdavsRunpod
AWSvsLambda
AWSvsRunpod
LambdavsVast
AWSvsVast
AWSvsGoogle Cloud
NebiusvsRunpod
LambdavsNebius
Google CloudvsNebius
AWSvsAzure
HyperstackvsLambda
RunpodvsShadeform
← Back to GPU Finder