Get alerts
All products/GPU cloud & model hosting/Runpod

Runpod pricing, price hikes and cheaper alternatives

Runpod (runpod.io) is a GPU cloud platform. Pay-as-you-go per GPU-hour for Pods and Serverless with no monthly plan; clusters and reserved capacity are partly quoted by sales.. No free plan.

What Runpod costs

  • 100 H100 hours: $199/mo (H100 PCIe 80 GB on Community Cloud at $1.99/hr x 100 h = $199. It is the cheapest 80 GB H100 listed (H100 SXM Community $2.69, H100 PCIe Secure $2.89). No plan fee or recurring credit applies. The page does not itemize regions (it says 30+), so US/EU availability is not verified.)
  • 24/7 GPU endpoint: $116.80/mo (RTX A5000 24 GB Pod on Community Cloud at $0.16/hr x 730 h = $116.80. It is the cheapest listed GPU with 24 GB or more. The page lists no separate CPU or RAM rate, so none is added. A 24 GB serverless worker at $0.69/hr would cost $503.70 and is not used. US/EU availability is not verified.)
  • 1,000 sandbox hours: not offered (n/a: the pricing page lists no sandbox product or isolated-sandbox SDK.)

Plan compared: Pay-as-you-go Pods, Community Cloud: 1x H100 PCIe 80 GB at $1.99/hr for 100 hours a month, with no plan fee or credit.. Prices are USD per month at on-demand, non-interruptible rates for three fixed workloads: 100 hours on one H100 80 GB GPU, one 24 GB-class GPU serving a model 24/7 (730 hours), and 1,000 hours of 2 vCPU / 4 GB agent sandboxes. They include required plan fees and separately billed CPU and RAM, net of recurring monthly credits; a dash means the product does not do that job.

What Runpod does

Core: GPU instances, Serverless GPU functions.

Important: H200 and Blackwell GPUs, Low-cost 24 GB GPUs, Multi-GPU nodes and clusters (partly), Dedicated model endpoints, Ready-to-run model APIs, Bring your own container, Persistent volumes, API, CLI and SDKs.

Also has: Public model and app hub (partly), Managed fine-tuning (partly), Hosted notebooks and IDEs (partly), Agent code sandboxes (partly), Sandbox snapshots and long sessions (partly), Batch jobs and queues (partly), Spot and reserved discounts (partly), Open GPU host marketplace (partly), Open source and own cloud (partly), Compliance and data residency (partly).

Runpod price hikes

Alternatives to Runpod

Runpod is the cheapest GPU cloud platform we track at 100 H100 hours a month. The others:

Runpod alternatives: 4 GPU cloud platforms compared, at every size.

Other kinds of products that cover part of Runpod

  • Vercel (Hosting): covers 44% of Runpod; missing or partial: GPU instances, Serverless GPU functions, H200 and Blackwell GPUs, Low-cost 24 GB GPUs, Multi-GPU nodes and clusters
  • Cloudflare Workers (Hosting): covers 41% of Runpod; missing or partial: GPU instances, Serverless GPU functions, H200 and Blackwell GPUs, Low-cost 24 GB GPUs, Multi-GPU nodes and clusters

Every change we logged

  • Oct 12, 2026 · Price increase: Secure Cloud Pod rates rose on six GPUs in one update: H100 SXM $3.49 to $3.99 per hour (+14.3%), B300 $7.89 to $8.99, H200 $4.59 to $5.29, B200 $6.79 to $7.99, A100 (variant not stated) $1.59 to $1.79, RTX 4090 $0.74 to $0.89.
  • Sep 7, 2026 · Price increase: Secure Cloud Pod rates rose on three GPUs: H100 SXM $3.29 to $3.49 per hour (+6.1%), A100 PCIe $1.39 to $1.59, L40S $0.99 to $1.09.
  • Aug 17, 2026 · Price increase: Secure Cloud Pod rates rose on five GPUs: H100 SXM $2.99 to $3.29 per hour (+10.0%), A100 SXM $1.49 to $1.59, B300 $7.39 to $7.89, H200 $4.39 to $4.59, B200 $5.89 to $6.79.
  • Jul 27, 2026 · Price increase: Secure Cloud RTX 3090 (24 GB) rose from $0.46 to $0.50 per hour (+8.7%).
  • Jul 13, 2026 · Price cut: Secure Cloud H100 SXM was cut from $3.29 to $2.99 per hour (-9.1%).
  • Jul 6, 2026 · Price increase: Serverless H200 rose from $5.58 to $5.93 per hour (+6.3%), while serverless RTX 6000 Pro fell from $4.00 to $3.49 (-12.8%) and L40, L40S and 6000 Ada fell from $1.90 to $1.75 (-7.9%).
  • Jun 30, 2026 · Price cut: Pod rate for L40 was cut from $0.99 to $0.82 per hour (-17.2%), while L40S rose from $0.86 to $0.99 (+15.1%).

Sources

Pricing page: runpod.io/pricing. Prices last checked Oct 16, 2026. Prices are from Runpod's official pricing page (updated 2026-10-08) and its GPU pricing page, which agree on the H100 PCIe, H100 SXM, RTX A5000, RTX 3090 and L4 rates. Changes are PricingSaaS diffs (tracking began 2026-04-29) with the conflicting rates left out, and pct uses H100 SXM Secure Cloud because the reference basis (H100 PCIe Community) has no sourced change history.