Inferensys

Glossary

WattTime API

A data service providing real-time marginal emissions rates for electrical grids globally, enabling automated carbon-aware scheduling by signaling the true carbon impact of consuming electricity at a specific moment.
Data scientist building training data pipeline on laptop, data preprocessing visible, technical workspace.
MARGINAL EMISSIONS DATA SERVICE

What is WattTime API?

A technical definition of the WattTime API, a data service providing real-time marginal emissions rates for electrical grids globally to enable automated carbon-aware scheduling.

The WattTime API is a data service that provides real-time, location-based marginal emissions rates for electrical grids, quantifying the carbon impact of consuming one additional megawatt-hour of electricity at a specific moment. It signals the true, dynamic carbon intensity of power by identifying the specific generating resource that will respond to a change in demand.

By integrating this API into workload orchestration platforms, systems can execute carbon-aware scheduling, automatically time-shifting compute tasks to periods of lower carbon intensity. This enables precise Scope 2 emissions reduction for cloud workloads without decreasing operational throughput, moving beyond static annual averages to dynamic, consequential accounting.

WATTTIME API

Key Features

Core capabilities of the WattTime API that enable automated, real-time carbon-aware decision-making for enterprise workloads.

01

Real-Time Marginal Emissions Signal

Provides a marginal emissions rate (MOER) in lbs CO2 per MWh for the specific grid balancing authority where electricity is being consumed. Unlike average grid data, the MOER signals the carbon intensity of the power plant that must ramp up or down to meet a change in demand, offering a causally accurate measure for dynamic compute scheduling.

5 min
Data Granularity
Global
Coverage
03

Automated Grid Region Detection

Automatically maps a workload's geographic coordinates or cloud provider region to the correct grid balancing authority. This removes the complexity of manually mapping IP addresses or data center locations to electrical grid operators, ensuring the correct marginal signal is applied without manual configuration.

04

Avoided Emissions Calculation

Quantifies the carbon impact of shifting workloads by comparing the emissions that would have occurred against what actually occurred. The API calculates avoided emissions by analyzing the counterfactual baseline, providing verifiable metrics for ESG reporting and carbon credit validation.

05

Programmatic Access for CI/CD Pipelines

Designed for integration into automated infrastructure. A simple RESTful JSON API with SDKs for Python and other languages allows DevOps teams to embed carbon signals directly into Kubernetes schedulers, CI/CD runners, and infrastructure-as-code tools, enabling GreenOps automation.

06

Third-Party Verified Methodology

The marginal emissions methodology is independently validated and aligns with the Greenhouse Gas Protocol's Scope 2 Guidance for location-based market instruments. This provides the audit-grade data integrity required for regulatory disclosures under frameworks like the EU CSRD.

WATTTIME API

Frequently Asked Questions

Essential questions about the WattTime API, a data service providing real-time marginal emissions rates for electrical grids globally, enabling automated carbon-aware scheduling.

The WattTime API is a data service that provides real-time marginal emissions rates for electrical grids globally. It works by analyzing grid data to determine the specific power plant that is responding to changes in demand at any given moment—the marginal generator—and calculating the carbon intensity of that plant's electricity. Unlike average grid emission factors, which smooth out the carbon impact over time, WattTime signals the true, dynamic carbon cost of consuming electricity at a specific location and moment. The API delivers this data programmatically, allowing software systems to query the current marginal operating emissions rate (MOER) and automatically shift flexible workloads to cleaner periods.

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.