GB200 rental price: from $8.00/hr today

GB200 offers: 2 configurations from 2 providers

On-demand and secure-cloud rates are capacity the provider runs, with no reservation. Spot and community capacity can be interrupted. A marketplace host is one independent owner's server listed on a marketplace. Prices are per GPU per hour, before storage and data transfer. Every row shows the date it was checked.

GB200 price history and range

GB200 prices by provider and billing type

Each provider's cheapest rate in each billing type, per GPU-hour. The provider name opens its GB200 page with node sizes, fees and its own history.

GB200 hardware specs

Manufacturer datasheet figures for the GB200. Source: www.nvidia.com.

LLM memory estimates for this GPU

These models have an estimated inference footprint within 186 GB at one or more precisions. The scenario uses 4,096 tokens, batch 1, a 16-bit KV cache and 20% runtime headroom. A memory fit does not establish speed or software compatibility.

Models that need more than one GB200

• Qwen3 Coder 480B A35B at 16-bit needs about 1154 GB with the same context and headroom, so at least 7 cards. No provider we track publishes a GB200 node of that size today. At 8 or 4 bit it needs less; the model page recalculates.
• GLM 4.5 at 16-bit needs about 862 GB with the same context and headroom, so at least 5 cards. No provider we track publishes a GB200 node of that size today. At 8 or 4 bit it needs less; the model page recalculates.
• GLM 4.6 at 16-bit needs about 858 GB with the same context and headroom, so at least 5 cards. No provider we track publishes a GB200 node of that size today. At 8 or 4 bit it needs less; the model page recalculates.

Compare model memory requirements or change context and memory assumptions on a model's page.

When to choose the GB200, and what to pick instead

Start from memory: your model, its context and your batch must fit in 186 GB (the model-fit table above estimates it). Then the GB200 is the right card unless one of these fits your job better.

• Pick the B200 (180 GB, from $3.49/hr) when your model fits in 180 GB and you can wait longer: it costs 56% less per hour.
• Pick the B300 (288 GB, from $4.99/hr) when the model or its cache does not fit in 186 GB.

Rent or buy the GB200

We have no confirmed GB200 purchase price from the stores we check; renting starts at $8.00 per GPU-hour. A seller quote entered in the calculator gives the break-even.

Compare GB200 purchase prices and quote availability, then enter your usage and electricity costs in the rent-versus-buy calculator. Purchase listings have their own variants, stock status and update dates. Compare two GPUs.

GB200 questions

How much does it cost to rent a GB200?

The cheapest GB200 rental we have is $8.00 per GPU-hour on GMI Cloud (on-demand), checked 22 October 2026.

How much VRAM does a GB200 have?

186 GB.

Where is the GB200 cheapest right now?

GMI Cloud has the cheapest checked on-demand rate today at $8.00 per GPU-hour.

GPUs near the GB200

• MI355X288 GB$5.99/hrcheapest on-demand
• MI350X288 GB$5.49/hrcheapest on-demand
• GB300288 GB$12.17/hrcheapest on-demand
• B200180 GB$3.49/hrcheapest on-demand
• B300288 GB$4.99/hrcheapest on-demand

Related

All 140 GB+ GPUs · GPUs under $2/hr · Price movers this week · GMI Cloud review