Blog/Cloud

Cloud Carbon Footprint: The open-source tool and its alternatives

How the open-source tool estimates cloud emissions, where it stops, and how AWS, Azure, Google Cloud and Greenpixie compare.

James HallWritten by James HallDanielle GeurtsReviewed by Danielle GeurtsUpdated 17 Nov 2026

GreenOps guide

Cloud Carbon Footprint and its alternatives compared

AWSAzureGoogle Cloud

Cloud Carbon Footprint, often shortened to CCF, is a free, open-source tool that estimates the energy consumption and carbon emissions of workloads on Amazon Web Services (AWS), Microsoft Azure and Google Cloud from billing and resource usage data. Thoughtworks, a technology consultancy, created it, and its repository dates from November 2020. It suits engineering teams at companies that want one published method across all three clouds and are happy to host and maintain the software themselves. The main alternatives are the free carbon reports that each cloud service provider offers for its own platform, and Greenpixie, which adds energy, carbon and water figures to every line of the cloud bill across AWS, Azure and Google Cloud.

Every detail about CCF and the provider reports comes from their own documentation and repositories, checked in November 2026.

Key takeaways

  • CCF is free under the Apache 2.0 licence and covers the three major cloud providers, AWS, Microsoft Azure and Google Cloud, with one published method.
  • It estimates electricity use from vCPU hours, processor wattage and utilisation, then applies a data centre efficiency factor and grid emission factors.
  • Commits paused between July 2024 and December 2025. The latest release came out on 23 April 2026, and the main repository has had no commits since.
  • The provider reports are free for one cloud each. Greenpixie suits companies that want line-item carbon, energy and water figures across all three without running their own software.

What is Cloud Carbon Footprint?

Cloud Carbon Footprint is an open source tool with a dashboard, an API and a command-line tool, built in separate modules so a team can run the whole app or only the parts it needs. It reads billing records from each provider, which means the AWS Cost and Usage Report through Amazon Athena, the Google Cloud billing export in BigQuery and the Azure Consumption Management API. Because it connects to multiple cloud providers at once, a team gets one view of its cloud footprint, running in its own environment. It turns that activity into estimated energy consumption in kilowatt-hours and carbon emissions in metric tons of CO2e.

Its documentation describes it as a way to measure, monitor and reduce carbon emissions from the cloud. CCF also returns right-sizing recommendations from AWS Cost Explorer and Compute Optimizer, Google Cloud Recommender and Microsoft's Azure Advisor, with the estimated energy and CO2e savings of each one.

Thoughtworks lists Cloud Carbon Footprint on its Technology Radar, where it last appeared in the September 2023 edition in the Trial ring. The Radar entry says that "settling on one tool, taking a baseline and improving from that baseline is the key usage scenario". CCF fits that use well, with one transparent method applied the same way every month to measure and track change over time.

How Cloud Carbon Footprint measures and estimates carbon emissions

The CCF methodology works up from usage in four steps:

  • Compute energy is vCPU hours multiplied by an average wattage between a minimum and maximum figure for each processor family, weighted by CPU utilisation. CCF uses the provider's resource usage data where it can get it and assumes 50% utilisation where it can't. Data storage, memory and networking get their own coefficients.
  • The result is multiplied by a fixed power usage effectiveness figure for each provider, which adds the cooling and power overhead of cloud data centres.
  • Energy is converted to carbon with location-based grid emission factors. Google Cloud uses Google's published regional figures, US regions on AWS and Azure use the EPA's eGRID data, and most other regions use country-level factors.
  • Embodied carbon from manufacturing the hardware is added for compute.

The processor wattage figures come from SPECpower benchmark results. They are now calculated with the ccf-coefficients tool, which replaced the archived cloud-carbon-coefficients notebook. Its last update, on 24 April 2026, added new SPECpower results and recent AWS and Google Cloud instance types.

That transparency is the main strength of Cloud Carbon Footprint. Because every coefficient and formula is published, a team can see exactly how each number was produced and change the assumptions it disagrees with.

Limits of Cloud Carbon Footprint's emissions data

The methodology page is open about where the estimates stop. The main limits are:

  • Estimates are single figures with no confidence range, and the documentation says they aren't a replacement for figures from the cloud providers.
  • The default emission factors are annual averages from before 2023, so they don't reflect the time of day. An Electricity Maps API key can replace them with live or historical data.
  • Embodied emissions are included for compute only, and networking covers traffic between data centres, not traffic to end users.
  • The 50% utilisation default applies wherever the provider doesn't supply utilisation data.
  • Only a subset of Microsoft Azure regions is supported.
  • There are no water figures, so the wider environmental impact isn't covered.

Running CCF also takes work, because a team has to host it, connect the billing export from each cloud platform and keep it up to date. On maintenance, the main repository shows no commits between July 2024 and December 2025, then a burst of work that updated dependencies, added new AWS usage types and processor data, and restored the recommendations API. That work ended with the 23 April 2026 release, and there have been no commits since. The repository isn't archived.

Cloud Carbon Footprint alternatives

Cloud provider carbon tools

The cloud giants each offer a free carbon report for their own cloud. All three report Scope 1, 2 and 3 emissions, publish monthly and use their own method, and each is designed mainly for carbon footprint reporting.

  • Amazon Web Services (AWS) logo
    AWS Sustainability console
  • Microsoft Azure logo
    Azure Carbon optimization
  • Google Cloud logo
    Google Cloud Carbon Footprint
  • AWS Sustainability console. Launched on 31 March 2026 and replaced the Customer Carbon Footprint Tool, which Amazon deprecated on 30 June 2026. It shows market-based and location-based figures by account, Region and service, publishes by the 21st of the following month and reports water withdrawals once a year. AWS states that its methodology has been verified by an independent third party.
  • Azure Carbon optimization. Tracks emissions down to individual resources, publishes by the 19th of the following month, keeps 12 months of data and suggests changes with their carbon and cost effect. Microsoft's older Emissions Impact Dashboard for Azure retires on 31 March 2027.
  • Google Cloud Carbon Footprint. Breaks emissions down by project, product and region, with hourly grid factors for location-based figures, and exports to BigQuery on the 15th of the following month. Per-customer energy is in private preview. Google's carbon footprint reporting methodology page notes that "these emissions reports have not yet been third-party verified or assured."

The provider reports need no hosting and give a reasonable annual figure for one cloud, which suits corporate sustainability reporting. Each one covers a single cloud provider with its own method, and the cloud service providers don't share one, so a multi-cloud estate ends up with reports that can't be added up like for like.

Greenpixie: carbon, energy and water on every billing line

Greenpixie is a data source for cloud carbon, energy and water figures, and works from the billing records a FinOps team already has. The team shares read-only access to them, such as the AWS Cost and Usage Report, and gets the same data back with energy, carbon and water columns on every line item. GPX Cloud has figures for millions of cloud usage types across AWS, Microsoft Azure and Google Cloud. Enrichment runs every day and re-enriches the month to date, so late billing adjustments are picked up.

Like Cloud Carbon Footprint, Greenpixie starts from energy consumption, breaking each SKU down into the compute, storage and networking hardware behind it. Electricity use is estimated from spec sheets, provider documentation and published research, and carbon and water are worked out with region-specific grid carbon intensity and water factors. Greenpixie maintains the model, so no one on the customer's team hosts software or updates coefficients. The methodology has been independently verified in accordance with ISO 14064-3, prepared with reference to ISO 14064-1, and designed to be in line with the GHG Protocol Corporate Standard. The full method is in the methodology documentation.

The data goes into the tools a team already uses, including FinOps platforms such as Flexera and Apptio, BI tools and the FOCUS framework. At Mastercard, Greenpixie data enabled $2M in cloud cost reductions, and Mastercard exceeded its cloud carbon reduction target.

Cloud Carbon Footprint and its alternatives compared

Cloud Carbon Footprint, the AWS, Azure and Google Cloud carbon tools, and Greenpixie compared
ToolCloud Carbon FootprintAWS Sustainability consoleAzure Carbon optimizationGoogle Cloud Carbon FootprintGreenpixie
Clouds coveredAWS, Azure, Google CloudAWSAzureGoogle CloudAWS, Azure, Google Cloud
Run bySelf-hosted, open-source contributorsAWSMicrosoftGoogleGreenpixie
Emission scopesOperational emissions, plus embodied emissions for computeScope 1, 2 and 3Scope 1, 2 and 3Scope 1, 2 and 3Location-based, including the cloud provider's Scope 3
Scope 2 methodLocation-basedMarket-based and location-basedNot stated in the docsMarket-based and location-basedLocation-based, with energy figures for applying other factors
Finest breakdownBilling data by account, service and regionAccount, Region and service, with EC2, S3 and CloudFront broken outSubscription, resource group and resourceBilling account, project, product and regionEvery billing line item, by SKU, region and workload
When data arrivesDepends on the billing exportMonthly, by the 21st of the following monthMonthly, by the 19th of the following monthMonthly, on the 15th of the following monthDaily, re-enriching the month to date
EnergyYes, estimatedNot reportedNot reportedIn private previewYes
WaterNot reportedAnnual water withdrawalsNot reportedNot reportedYes
CostFree, self-hostedFreeFreeFree, BigQuery fees for exportsFree to start in the app, with paid plans
Details in this table were checked against each tool's documentation and repository in November 2026. The tools change often, so check the linked sources before relying on a single detail.

How to reduce your cloud carbon footprint

Companies first need to measure the footprint of their digital infrastructure. Most of the changes that cut cloud emissions also cut costs, such as removing cloud waste like idle resources and right-sizing over-provisioned instances, because both reduce energy consumption across the cloud infrastructure. Grid carbon intensity also differs between regions, because some grids have more low-carbon electricity from hydro and other renewable energy, so where a flexible workload runs changes its carbon impact and its CO2 emissions. CCF's recommendations and Azure Carbon optimization show the estimated carbon saving of right-sizing and idle-resource changes, and Greenpixie puts carbon next to cost on the same billing data, so a team can see which savings cut both. Our guide to cloud cost optimisation covers the steps in more detail.

Which option fits

  • A team on one cloud that needs an annual figure can start with that provider's free report.
  • An engineering or technology team that wants one open method across clouds, and has time to host and maintain it, can use CCF to set a baseline.
  • A multi-cloud team that wants carbon in FinOps and engineering decisions needs line-item emissions data from one method, which is what Greenpixie provides.

Our guides on GreenOps metrics and KPIs and sustainable cloud computing cover what to do with the data, and Scope 3 cloud emissions covers the reporting side.

Get started for free and see the carbon, energy and water behind your own cloud bill.

FAQs

What is cloud carbon footprint?

A cloud carbon footprint is the greenhouse gas emissions caused by an organisation's use of cloud services and other digital infrastructure, measured in kilograms or tonnes of CO2e. It includes the electricity used to run the servers, data storage and networks behind each service, and the embodied emissions from manufacturing that hardware and the cloud data centres. Under the Greenhouse Gas Protocol, the provider counts backup generator fuel as Scope 1 direct emissions and purchased electricity as Scope 2 indirect emissions. For the cloud customer, all of it counts as Scope 3 emissions. Cloud Carbon Footprint is also the name of the open-source tool this page covers.

Is the open source Cloud Carbon Footprint tool still maintained?

The GitHub repository isn't archived, and after no commits between July 2024 and December 2025, work restarted and led to a release on 23 April 2026, the first since May 2024. There have been no commits since that release, as of 17 November 2026.

What methodology is used to calculate cloud carbon footprint?

Most methods estimate energy usage from cloud activity, multiply it by a data centre efficiency factor and convert it to carbon with grid emission factors, then add embodied emissions from hardware. CCF does this from vCPU hours and processor wattage. The provider reports allocate each provider's own emissions to customers by usage. Greenpixie models energy consumption for each billing SKU from the hardware behind it.

Why do cloud providers' carbon tools give different numbers?

Each provider draws its own boundary and uses its own method. AWS includes the embodied carbon of data centre buildings in Scope 3, while Microsoft's methodology doesn't include buildings yet. Scope 2 emissions also depend on whether a figure is market-based or location-based, and Google uses hourly grid factors for its location-based figures. Each provider allocates its total emissions to customers in its own way, so the same workload can show different numbers on different clouds.

Is there a free cloud carbon footprint tool?

Cloud Carbon Footprint is free open-source software that a team hosts itself. The AWS Sustainability console, Azure Carbon optimization and Google Cloud Carbon Footprint are free for their own customers. Greenpixie can be tried for free in the app, where free tools show carbon, energy and water per pricing unit for AWS, Microsoft Azure and Google Cloud services. Some features, such as AI model data, are part of paid plans.

Measure your cloud emissions

See the energy, carbon and water behind your cloud bill

GPX Cloud adds energy, carbon and water data to AWS, Azure and Google Cloud billing data, line by line.

Sources

  1. Cloud Carbon Footprint: Project website (accessed November 2026)
  2. Cloud Carbon Footprint: Recommendations documentation (accessed November 2026)
  3. Thoughtworks: Technology Radar, Cloud Carbon Footprint (Trial, September 2023)
  4. Cloud Carbon Footprint: Methodology documentation (accessed November 2026)
  5. Cloud Carbon Footprint: ccf-coefficients repository, GitHub (last updated 24 April 2026)
  6. Cloud Carbon Footprint: cloud-carbon-footprint repository, GitHub (accessed 17 November 2026)
  7. Cloud Carbon Footprint: Releases, GitHub (latest release 23 April 2026)
  8. AWS: AWS Sustainability console user guide (accessed November 2026)
  9. Microsoft Learn: Azure Carbon optimization overview (accessed November 2026)
  10. Google Cloud: Carbon Footprint reporting methodology (accessed November 2026)
  11. Greenpixie: Cloud methodology documentation (accessed November 2026)
James Hall
Written by
James Hall
Head of Tokenomics and GreenOps, Greenpixie
Danielle Geurts
Reviewed by
Danielle Geurts
GTM and Operations, Greenpixie
Share this guideLinkedInX