P2P marketplaces Live data

Meshive GPU rental provider logo Meshive.

GPU marketplace of verified hosts worldwide, billed by the minute with no minimums: on-demand and spot rates per GPU on RTX 5090 and RTX PRO 5000 Blackwell pods with templates and persistent storage, plus serverless model endpoints — prepaid credits, no egress fees.

At a glance

Business model
P2P marketplaces
Pricing model
On demand

When to pick Meshive

Best for

  • Cost-sensitive experimentation and hobbyist training.
  • Batch inference where occasional host swaps are fine.
  • Anyone willing to filter hosts by reliability score.

Avoid for

  • Production inference with strict latency SLAs.
  • Multi-host distributed training — hosts don't share infrastructure.
  • Workloads with data-residency or compliance requirements.
Price history

Daily median on Meshive's top GPUs.

Loading...

GPUs available on Meshive

GPU $/hr
Spot from $0.43/hr per GPU · interruptible
$0.71/hr Compare →

Compare with peers

Other providers of the same type. Open one to see its own prices.

Frequently asked

How is Meshive billing measured?
Meshive bills per minute, as its own pricing page states. Headline $/hr is the comparison unit across providers.
Which regions does Meshive offer GPUs in?
We haven't confirmed where Meshive runs GPUs. Its site lists the locations you can pick when you deploy.
Does Meshive offer an SLA?
Hyperscalers (AWS, GCP, Azure, Oracle) publish formal SLAs. First-party clouds (Lambda, CoreWeave) offer support contracts. P2P marketplaces and decentralized networks have no SLA — uptime depends on the individual host.
What's the refund / cancellation policy on Meshive?
Meshive bills per minute, so stopping the instance stops the meter at the end of the current increment. Pre-paid credits and committed-use discounts have their own terms; check Meshive's billing docs before pre-paying.