The operating system for AI compute

Describe your
workload.

Tell BOOKCOMPUTE.com what you are building. Compare sourced public GPU rates and request confirmation of capacity and final terms from the Beta procurement desk.

AI
⌘ / Ctrl + Enter to advise
Try:
H100 SXM5/BetaQuote requiredSource review|B200 NVL72/BetaQuote requiredSource review|RTX PRO 6000/BetaQuote requiredSource review|TPU v5p/BetaQuote requiredSource review|H200 SXM/BetaQuote requiredSource review|A100 80GB/BetaQuote requiredSource review|MI300X/BetaQuote requiredSource review|H100 SXM5/BetaQuote requiredSource review|B200 NVL72/BetaQuote requiredSource review|RTX PRO 6000/BetaQuote requiredSource review|TPU v5p/BetaQuote requiredSource review|H200 SXM/BetaQuote requiredSource review|A100 80GB/BetaQuote requiredSource review|MI300X/BetaQuote requiredSource review|
Live Market

Reviewed GPU rates. Clear sources.

Beta · public pricing
NVIDIA · Hopper
H100 SXM5
Public rate
$3.99/gpu/hr
official public pricing · Source: Lambda self-serve pricing
Last updated: 2026-10-19 · Next review: 2026-10-26
Manual review of official public pricing, USD per GPU-hour · Confidence: 90/100
NVIDIA · Blackwell
B200 NVL72
Public rate
Quote/gpu/hr
No reviewed public rate yet.
NVIDIA · Ada
RTX PRO 6000
Public rate
Quote/gpu/hr
No reviewed public rate yet.
Google · Google TPU
TPU v5p
Public rate
Quote/gpu/hr
No reviewed public rate yet.
NVIDIA · Hopper
H200 SXM
Public rate
Quote/gpu/hr
No reviewed public rate yet.
NVIDIA · Ampere
A100 80GB
Public rate
Quote/gpu/hr
No reviewed public rate yet.
AMD · AMD Instinct
MI300X
Public rate
Quote/gpu/hr
No reviewed public rate yet.
Provider Marketplace

Beta providers. One procurement surface.

Explore real providers and sourced public GPU rates. Request confirmation of capacity, regions and commercial terms.

View directory →
CodeProviderInventoryUptimeFrom
  • AWS
    Amazon Web Services
    Hyperscaler · Directory
    —
    —
    View →
  • AZUR
    Microsoft Azure
    Hyperscaler · Directory
    —
    —
    View →
  • GCP
    Google Cloud
    Hyperscaler · Directory
    —
    —
    View →
  • CORE
    CoreWeave
    Tier 1 · Directory
    —
    —
    View →
  • LAMB
    Lambda
    Tier 1 · Directory
    —
    —
    View →
  • RUNP
    RunPod
    Neocloud · Directory
    —
    —
    View →
  • VAST
    Vast.ai
    Spot · Directory
    —
    —
    View →
  • NEBI
    Nebius
    Tier 1 · Directory
    —
    —
    View →
  • FLUI
    Fluidstack
    Tier 1 · Directory
    —
    —
    View →
  • CRUS
    Crusoe
    Tier 1 · Directory
    —
    —
    View →
  • TOGE
    Together AI
    Tier 1 · Directory
    —
    —
    View →
  • GENE
    Genesis Cloud
    Neocloud · Directory
    —
    —
    View →
  • TENS
    TensorDock
    Neocloud · Directory
    —
    —
    View →
  • HYPE
    Hyperstack
    Neocloud · Directory
    —
    —
    View →
  • OVHC
    OVHcloud
    Tier 1 · Directory
    —
    —
    View →
  • PHOE
    PhoenixNAP
    Neocloud · Directory
    —
    —
    View →
  • DATA
    DataCrunch
    Neocloud · Directory
    —
    —
    View →
  • LATI
    Latitude.sh
    Neocloud · Directory
    —
    —
    View →
  • EXOS
    Exoscale
    Neocloud · Directory
    —
    —
    View →
  • PAPE
    Paperspace (DigitalOcean)
    Neocloud · Directory
    —
    —
    View →
Why BookCompute

An institutional layer for a fragmented market.

01

One Brief

Capture your accelerator, quantity, region and duration requirements.

02

Provider Directory

Real provider listings. A listing does not imply a partnership or endorsement.

03

Sourced Prices

Reviewed public rates with source links and dates, not simulated live feeds.

04

Private Requests

Save structured requests to your account for Beta review. No guaranteed response or capacity.

05

Trust First

Unconfirmed capacity and compliance stay unconfirmed until evidence is available.

Enterprise RFQ

Request 500 GPUs.
Receive private bids.

Save a structured request for Beta review. Response times, pricing, topology and capacity require confirmation.

Request requirementsStatusDetails
  • Accelerator and quantity—Required
  • Region and duration—Required
  • Budget and timeline—Required
Spot Marketplace

Last-minute compute.

Beta
No confirmed spot offers
Beta
—
Unconfirmed
Price Intelligence

Every SKU, every venue, every hour.

Historical series require dated source evidence. No verified history or market forecast is published in this Beta.

30D spread
Not measured
Best venue
Not ranked
Forecast H100
Not published
Advisory
Confirm quotes
90-Day price history · $/GPU/hr
No verified price history yet.
H100 SXM5A100 80GBB200 (fwd)
Benchmark Center

Compare compute like an analyst.

No measured provider benchmark results are published yet. Hardware specifications are not workload benchmarks.

ProviderChipsetLatency P99Training TF/sInference tok/sInterconnect$/hr
No verified benchmark measurements yet.
0
Provider directory
Beta
Product stage
Sources
Pricing evidence
Private
Request visibility
Ready?

Optimize your AI infrastructure.

Compare sourced public rates and submit your requirements. Join the Beta with a real compute brief.