GitHub delivery evidence
Build and verify the change with supported GitHub context.
AWS · GCP · AzureFacetedGPU specs
Compare compute without twelve open tabs.
One normalized catalog across AWS, GCP, and Azure. Filter by vCPU, memory, GPU, and purchase option, sort by price per resource, and read the GPU specs in place, then confirm the final number with the provider.
- 4 vCPU · 16 GB · aws$0.17/hr
- 4 vCPU · 16 GB · gcp$0.19/hr
- 4 vCPU · 16 GB · azure$0.21/hr
Compute pricing is planning input for the engineering stage.
One connected loop, held on the stage this capability serves. The other stages stay as context so you can see what feeds in and what comes next.
BD
Discovery
Feedback
Product
Engineering
Release
Reliability
Evidence
Engineering
Build and verify the change.
Available now
VCS insights
Inspect delivery workflow activity in the connected release workspace.
Test intelligence
Investigate test outcomes and flakiness evidence across runs.
API Explorer
Keep supported API contracts inspectable alongside release evidence.
Compute pricing
Compare public cloud compute pricing as planning evidence.
Where the monthly estimate goes.
Estimated monthly spend by service from on-demand catalog prices. The bars fill to each figure while the total counts up, so the widest bar shows where the money actually lands. Confirm the final number with the provider.
estimated monthly · on-demand~$6840/mo
- GPU training · 4x H100~$4,800/mo
- API compute · 6x 8 vCPU~$1,180/mo
- Batch jobs · spot~$540/mo
- Staging · 3x 4 vCPU~$320/mo
Estimate, trend, and unit cost.
The trend, honestly dashed.
The actual monthly line draws in solid; the projection past now is dashed, so nothing pretends to be measured. A rising slope is a planning signal, not a bill.
actual solidprojection dashedestimate
projected spend
actual projectedFebnowproj. Aug
Cost per deploy and per env.
Derived figures that count up to their estimated value, so a monthly number becomes a per-deploy and per-environment cost you can reason about. Every tile stays marked estimate.
per deployper environmentblended rate
Cost / deploy
~$42
monthly ÷ deploys
Cost / environment
~$1140/mo
across 6 envs
Cost / vCPU-hour
~$0.043
blended on-demand
Cost / GPU-hour
~$3.9
H100 on-demand
One instance, three providers.
Every offer is normalized to the same row so a like-for-like compare is one glance. On-demand, spot, and reserved sit together per instance and region.
same shapeon-demand + spotper region
- 8 vCPU · 32 GB · on-demand$0.34/hr
- 8 vCPU · 32 GB · spot$0.11/hr
- 8 vCPU · 32 GB · reserved$0.22/hr
Faceted, not flat.
Providers, purchase options, operating systems, and GPU vendors are all filterable facets, so a comparison narrows fast.
Providers
Normalized together.
awsgcpazure
Purchase options
How you buy.
on_demandspotreservedcommitted_use
Operating systems
Priced per OS.
linuxwindowsrhelslesubuntu_pro
GPU vendors
Curated specs.
nvidiaamdawsgoogleintel
Architecture
CPU families.
x86_64arm64
Sort by
Value ratios.
hourlyPricepricePerVcpupricePerMemoryGbgpuMemoryGb
An honest limit
This is a normalized reference index, not a purchasing path. Rows carry an indexed and price-effective date, GPU specs are curated per model rather than live from the feed, and final purchasing decisions belong with the provider.
3
providers normalized into one comparable catalog
Est.
figures only, confirm the final number with the provider
Spot
and on-demand pricing carried per instance and region
Per-row
price-effective and indexed timestamps on every row
See what else the engineering stage carries.
Start your infrastructure comparison in one place.
Start free
A faster first pass, then confirm with the provider.