GPU Rental Prices
SearchCtrl K

Sources and update dates

Rental prices come from providers, hardware specifications from manufacturers, model parameters from publisher records, and purchase prices from named retailers. Competitor price tables are not sources.

GPU rental prices

Checked dates belong to each provider. A recent page build does not renew an older observation. Rates are USD per GPU-hour, before taxes; a quoted node total is divided by its published GPU count.

ProviderOfficial pricing sourceCheckedPrice check
Daytonabilling.app.daytona.io21 November 2026Prices read
Liumlium.io21 November 2026Prices read
Impossible Cloudwww.impossiblecloud.com21 November 2026Prices read
Synteqsynteq.com21 November 2026Prices read
HyperAIhyper.ai21 November 2026Prices read
Runcratewww.runcrate.ai21 November 2026Prices read
GPUwerkgpuwerk.com21 November 2026Prices read
RunPodwww.runpod.io21 November 2026Prices read
Lambdalambda.ai21 November 2026Prices read
Hyperstackwww.hyperstack.cloud21 November 2026Prices read
CoreWeavewww.coreweave.com21 November 2026Prices read
Nebiusdocs.nebius.com21 November 2026Prices read
Verda (DataCrunch)datacrunch.io21 November 2026Prices read
Voltage Parksupport.voltagepark.com21 November 2026Prices read
Crusoewww.crusoe.ai21 November 2026Prices read
Jarvislabsjarvislabs.ai21 November 2026Prices read
OVHcloudwww.ovhcloud.com21 November 2026Prices read
Spheronwww.spheron.network21 November 2026Prices read
Microsoft Azureprices.azure.com21 November 2026Prices read
AWSb0.p.awsstatic.com21 November 2026Prices read
Oracle Cloudapexapps.oracle.com21 November 2026Prices read
Hetznerwww.hetzner.com21 November 2026Prices read
Together AIwww.together.ai21 November 2026Prices read
falfal.ai21 November 2026Prices read
Modalmodal.com21 November 2026Prices read
Vultrwww.vultr.com21 November 2026Prices read
Massed Computemassedcompute.com21 November 2026Prices read
TensorDockwww.tensordock.com21 November 2026Prices read
SaladCloudsalad.com21 November 2026Prices read
Replicatereplicate.com21 November 2026Prices read
Fireworks AIfireworks.ai21 November 2026Prices read
Basetenwww.baseten.co21 November 2026Prices read
DigitalOceanwww.digitalocean.com21 November 2026Prices read
GMI Cloudwww.gmicloud.ai21 November 2026Prices read
Prime Intellectwww.primeintellect.ai31 October 2026Earlier prices kept, with their dates; the latest price check failed
Cerebriumwww.cerebrium.ai21 November 2026Prices read
Koyebwww.koyeb.com21 November 2026Prices read
Thunder Computewww.thundercompute.com21 November 2026Prices read
Latitude.shwww.latitude.sh21 November 2026Prices read
Hot Aislehotaisle.xyz21 November 2026Prices read
Scalewaywww.scaleway.com21 November 2026Prices read
Google Cloudcloud.google.com21 November 2026Prices read

Availability is reported only when the source explicitly publishes stock. “Not reported” means there is no stock observation, even when the price is current. Reserved rates carry a term, minimum volume or an explicit statement that the commitment was not published. Monthly rates use 720 hours; annual Latitude plans use 8,760 hours. Those conversions do not establish payment timing or cancellation rights.

Retired price sources

SF Compute public price history ends with the last public observation on 24 September 2026. Prices have not been public since 25 September 2026. Collection ended on 4 October 2026; this source is excluded from current provider counts.

Hardware specifications

Specifications checked on 18 July 2026, unless a later date is given. Manufacturer specifications describe the named variant, not measured application performance. A peak compute figure is not a benchmark. Unspecified attributes are left out.

Company facts

Rubin sources and specifications

Specifications checked on 5 October 2026. GPU figures refer to NVIDIA's individual-GPU table, not a cloud instance or a complete rack. Peak compute specifications are not measured workload performance.

ClaimOfficial sourceQuoted evidence
Rubin individual GPU specificationswww.nvidia.com288 GB HBM4; 22 TB/s; 3.6 TB/s; 4 PFLOPS; 50 PFLOPS
Vera Rubin NVL72www.nvidia.com72 NVIDIA Rubin GPUs
B200 GPU memorydocs.nvidia.comB200 180 GB product
B300 GPU memorydocs.nvidia.com8 x 288 GB = 2.3 TB total
H200 SXM GPU memorywww.nvidia.com141GB
B200 memory typewww.nvidia.com1,440 GB total, 64 TB/s HBM3e bandwidth
B300 memory typewww.nvidia.comHBM3E
NVIDIA launch and cloud partners, 5 January 2026nvidianews.nvidia.comCoreWeave, Lambda, Nebius and Nscale.
CoreWeave production announcementwww.coreweave.comannounced the availability of NVIDIA Vera Rubin NVL72 on CoreWeave
Lambda Rubin product pagelambda.aiAvailable H2 2026.
Spheron pre-order listingwww.spheron.networkShips H2 2026

NVIDIA publishes two sets of per-GPU Rubin figures. Its HGX table lists 22 TB/s of memory bandwidth and 3.6 TB/s of NVLink bandwidth; its Vera Rubin NVL72 table lists 19.2 TB/s and 3 TB/s, for an at-scale AI factory. Ask the provider which system it rents before comparing performance.

Provider fees, access and compliance

Checked 7 July 2026; cells rechecked later show their own date on the fees and enterprise matrices. These are documentation observations, not an audit of a provider. Conditional rates and unpublished fields cannot be turned into a complete bill.

Model parameter counts and memory

Parameter counts come from the publisher’s Safetensors metadata. For standard attention, cache bytes = 2 × layers × key/value heads × head dimension × context tokens × batch size × 2 bytes. Weight bytes = parameters × precision bits ÷ 8. Memory tables use decimal GB, with editable overhead assumptions; they do not guarantee that a particular runtime will load a model.

The full-training scenario uses the Transformers mixed-precision AdamW memory accounting: 18 bytes per parameter, plus activations supplied by the reader. Adapter training applies that accounting only to the assumed trainable parameter fraction. These are budget estimates, not throughput measurements.

ModelPublisherCheckedCache estimate
Llama 3.1 8Bmeta-llama/Llama-3.1-8B-Instruct21 November 2026Configuration restricted or attention type not modelled
Llama 3.1 70Bmeta-llama/Llama-3.1-70B-Instruct21 November 2026Configuration restricted or attention type not modelled
Llama 3.1 405Bmeta-llama/Llama-3.1-405B-Instruct21 November 2026Configuration restricted or attention type not modelled
Llama 3.2 1Bmeta-llama/Llama-3.2-1B-Instruct21 November 2026Configuration restricted or attention type not modelled
Llama 3.2 3Bmeta-llama/Llama-3.2-3B-Instruct21 November 2026Configuration restricted or attention type not modelled
Llama 3.3 70Bmeta-llama/Llama-3.3-70B-Instruct21 November 2026Configuration restricted or attention type not modelled
Qwen2.5 0.5BQwen/Qwen2.5-0.5B-Instruct21 November 2026Published configuration
Qwen2.5 1.5BQwen/Qwen2.5-1.5B-Instruct21 November 2026Published configuration
Qwen2.5 3BQwen/Qwen2.5-3B-Instruct21 November 2026Published configuration
Qwen2.5 7BQwen/Qwen2.5-7B-Instruct21 November 2026Published configuration
Qwen2.5 14BQwen/Qwen2.5-14B-Instruct21 November 2026Published configuration
Qwen2.5 32BQwen/Qwen2.5-32B-Instruct21 November 2026Published configuration
Qwen2.5 72BQwen/Qwen2.5-72B-Instruct21 November 2026Published configuration
Qwen2.5 Coder 0.5BQwen/Qwen2.5-Coder-0.5B-Instruct21 November 2026Published configuration
Qwen2.5 Coder 1.5BQwen/Qwen2.5-Coder-1.5B-Instruct21 November 2026Published configuration
Qwen2.5 Coder 3BQwen/Qwen2.5-Coder-3B-Instruct21 November 2026Published configuration
Qwen2.5 Coder 7BQwen/Qwen2.5-Coder-7B-Instruct21 November 2026Published configuration
Qwen2.5 Coder 14BQwen/Qwen2.5-Coder-14B-Instruct21 November 2026Published configuration
Qwen2.5 Coder 32BQwen/Qwen2.5-Coder-32B-Instruct21 November 2026Published configuration
Qwen3 0.6BQwen/Qwen3-0.6B21 November 2026Published configuration
Qwen3 1.7BQwen/Qwen3-1.7B21 November 2026Published configuration
Qwen3 4BQwen/Qwen3-4B21 November 2026Published configuration
Qwen3 8BQwen/Qwen3-8B21 November 2026Published configuration
Qwen3 14BQwen/Qwen3-14B21 November 2026Published configuration
Qwen3 32BQwen/Qwen3-32B21 November 2026Published configuration
Qwen3 30B A3BQwen/Qwen3-30B-A3B21 November 2026Published configuration
Qwen3 235B A22BQwen/Qwen3-235B-A22B21 November 2026Published configuration
DeepSeek R1 Distill Qwen 1.5Bdeepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B21 November 2026Published configuration
DeepSeek R1 Distill Qwen 7Bdeepseek-ai/DeepSeek-R1-Distill-Qwen-7B21 November 2026Published configuration
DeepSeek R1 Distill Qwen 14Bdeepseek-ai/DeepSeek-R1-Distill-Qwen-14B21 November 2026Published configuration
DeepSeek R1 Distill Qwen 32Bdeepseek-ai/DeepSeek-R1-Distill-Qwen-32B21 November 2026Published configuration
DeepSeek R1 Distill Llama 8Bdeepseek-ai/DeepSeek-R1-Distill-Llama-8B21 November 2026Published configuration
DeepSeek R1 Distill Llama 70Bdeepseek-ai/DeepSeek-R1-Distill-Llama-70B21 November 2026Published configuration
DeepSeek V3deepseek-ai/DeepSeek-V321 November 2026Configuration restricted or attention type not modelled
DeepSeek R1deepseek-ai/DeepSeek-R121 November 2026Configuration restricted or attention type not modelled
Mistral 7B v0.3mistralai/Mistral-7B-Instruct-v0.321 November 2026Published configuration
Mixtral 8x7B v0.1mistralai/Mixtral-8x7B-Instruct-v0.121 November 2026Published configuration
Mixtral 8x22B v0.1mistralai/Mixtral-8x22B-Instruct-v0.121 November 2026Published configuration
Mistral Small 24B 2501mistralai/Mistral-Small-24B-Instruct-250121 November 2026Published configuration
Mistral Nemo 2407mistralai/Mistral-Nemo-Instruct-240721 November 2026Published configuration
Phi 3 mini 4kmicrosoft/Phi-3-mini-4k-instruct21 November 2026Configuration restricted or attention type not modelled
Phi 3 medium 4kmicrosoft/Phi-3-medium-4k-instruct21 November 2026Configuration restricted or attention type not modelled
phi 4microsoft/phi-421 November 2026Published configuration
Phi 4 minimicrosoft/Phi-4-mini-instruct21 November 2026Configuration restricted or attention type not modelled
gemma 2 2bgoogle/gemma-2-2b-it21 November 2026Configuration restricted or attention type not modelled
gemma 2 9bgoogle/gemma-2-9b-it21 November 2026Configuration restricted or attention type not modelled
gemma 2 27bgoogle/gemma-2-27b-it21 November 2026Configuration restricted or attention type not modelled
gemma 3 1bgoogle/gemma-3-1b-it21 November 2026Configuration restricted or attention type not modelled
gemma 3 4bgoogle/gemma-3-4b-it21 November 2026Configuration restricted or attention type not modelled
gemma 3 12bgoogle/gemma-3-12b-it21 November 2026Configuration restricted or attention type not modelled
gemma 3 27bgoogle/gemma-3-27b-it21 November 2026Configuration restricted or attention type not modelled
Llama 4 Scout 17B 16Emeta-llama/Llama-4-Scout-17B-16E-Instruct21 November 2026Configuration restricted or attention type not modelled
Llama 4 Maverick 17B 128Emeta-llama/Llama-4-Maverick-17B-128E-Instruct21 November 2026Configuration restricted or attention type not modelled
Qwen3 4B 2507Qwen/Qwen3-4B-Instruct-250721 November 2026Published configuration
Qwen3 30B A3B 2507Qwen/Qwen3-30B-A3B-Instruct-250721 November 2026Published configuration
Qwen3 235B A22B 2507Qwen/Qwen3-235B-A22B-Instruct-250721 November 2026Published configuration
Qwen3 Coder 30B A3BQwen/Qwen3-Coder-30B-A3B-Instruct21 November 2026Published configuration
Qwen3 Coder 480B A35BQwen/Qwen3-Coder-480B-A35B-Instruct21 November 2026Published configuration
Mistral Small 3.2 24B 2506mistralai/Mistral-Small-3.2-24B-Instruct-250621 November 2026Published configuration
GLM 4.5 Airzai-org/GLM-4.5-Air21 November 2026Published configuration
GLM 4.5zai-org/GLM-4.521 November 2026Published configuration
Phi 4 reasoningmicrosoft/Phi-4-reasoning21 November 2026Published configuration
gpt-oss-20bopenai/gpt-oss-20b21 November 2026Configuration restricted or attention type not modelled
gpt-oss-120bopenai/gpt-oss-120b21 November 2026Configuration restricted or attention type not modelled
Kimi K2moonshotai/Kimi-K2-Instruct21 November 2026Configuration restricted or attention type not modelled
DeepSeek V3.1deepseek-ai/DeepSeek-V3.121 November 2026Configuration restricted or attention type not modelled
DeepSeek R1 0528deepseek-ai/DeepSeek-R1-052821 November 2026Configuration restricted or attention type not modelled
Kimi K2 0905moonshotai/Kimi-K2-Instruct-090521 November 2026Configuration restricted or attention type not modelled
GLM 4.6zai-org/GLM-4.621 November 2026Published configuration
Qwen3 Next 80B A3BQwen/Qwen3-Next-80B-A3B-Instruct21 November 2026Configuration restricted or attention type not modelled
QwQ 32BQwen/QwQ-32B21 November 2026Published configuration
Devstral Small 2507mistralai/Devstral-Small-250721 November 2026Published configuration
Ministral 8B 2410mistralai/Ministral-8B-Instruct-241021 November 2026Configuration restricted or attention type not modelled
Phi 4 reasoning plusmicrosoft/Phi-4-reasoning-plus21 November 2026Published configuration

LLM API rates

Serverless billing

Runpod billing documentation. Checked 18 October 2026. Startup, execution and idle timeout are billed, with whole-second rounding; active-worker discounts require a sales inquiry.

Purchase prices

A quote-only listing does not establish a purchase price. Expired or unconfirmed prices are excluded, not assigned a new date.

B200, quote only: Lenovo ThinkSystem NVIDIA HGX B200 180GB 1000W GPU from Eton Technology. Checked 27 September 2026. NVIDIA B200 Tensor Core GPU from Eton Technology. Checked 27 September 2026. Supermicro HGX B200 8U Air-Cooled System from Eton Technology. Checked 27 September 2026. Supermicro HGX B200 4U Liquid-Cooled System from Eton Technology. Checked 27 September 2026. GIGABYTE AI-DLC-Rack NVIDIA-HGX-B200 from Eton Technology. Checked 27 September 2026. GIGABYTE AI-DLC-POD NVIDIA-HGX-B200 from Eton Technology. Checked 27 September 2026. GIGABYTE AI-AC-POD NVIDIA-HGX-B200 from Eton Technology. Checked 27 September 2026. Supermicro SRS-48UAC-B200SX from Eton Technology. Checked 27 September 2026.

GB200, quote only: Supermicro SRS-GB200-NVL72 from Eton Technology. Checked 27 September 2026.

Credits

Price history and corrections

History starts with the first dated observation. Daily records are append-only; identified errors are excluded or reclassified through a separate correction record. Statistical pages state whether they compare the same provider and configuration or a changing market sample.

Read the calculation methods. To report a price, use its row’s email link; it includes the provider, GPU and update date. Nothing submitted through that link is stored by this site.

Data attribution and reuse

Today's rental prices and the full daily history are offered under CC BY 4.0, with attribution to GPU Rental Prices. A fixed copy of the record's first two weeks (5 to 18 July 2026) has the identifier 10.5281/zenodo.21435394.

Compare your offers

Saved prices keep their original dates. Confirm the configuration, terms and availability at the source.

Select up to three offers to compare their rates and terms.

Type a GPU, a provider, a model or a question.