Honest, market‑leading pricing
Rent high-performance NVIDIA GPUs, billed in one-minute intervals — no minimums, no monthly commitments, no egress fees. Marketplace rates run up to 70% below traditional clouds.
* Rates shown are per-GPU starting prices (USD). Live pricing varies by availability and region in the console catalog.
One-minute intervals
Fees are calculated every 60 seconds based on the pods you have running. Stop a pod and billing stops with it.
No commitments
No minimums, no reservations, no monthly contracts. Top up credits and pay only for what you run.
No egress fees
Move your datasets and results in and out freely — bandwidth is never metered against you.
$1,362 /mo
Full control of the machine
Occupy the whole GPU over SSH and containers. Ideal for training, fine-tuning, and development — choose uninterrupted On-demand or Spot at up to 40% off.
Fully managed endpoint
Set a price cap and Meshive picks the cheapest GPU that fits. OpenAI-compatible API and autoscaling — billed only for the GPU time you use.
Pricing FAQ
How does GPU billing work on Meshive?
Meshive bills pod runtime in fixed one-minute intervals. Fees are calculated every 60 seconds based on the pods you have running and deducted from your credit balance. There are no minimums and no monthly commitments.
How much does a cloud RTX 5090 cost?
On-demand RTX 5090 pricing starts at the live rate shown above (around $0.50 per GPU-hour). Actual marketplace rates vary with availability and region, and rates are reviewed regularly.
Which GPUs can I rent on Meshive?
Meshive supports 40+ NVIDIA models — from consumer RTX 30/40/50-series cards to data-center GPUs like the A100, H100, H200, and B200. Live availability and spot prices for every SKU are in the console catalog.
How is Serverless inference priced?
Serverless endpoints let you set a price cap, and Meshive picks the cheapest GPU that fits — billed by the GPU-hour. See the Serverless docs for details.
For serverless inference, you set a price cap and Meshive picks the cheapest GPU that fits — see Pricing & Instances in the docs and the Serverless page.