Web Machine Learning Community Group

To make Machine Learning a first-class web citizen by incubating and developing a dedicated low-level Web API for machine learning inference in the browser.

Visit Website

At a Glance

MIT Computer Science and Artificial Intelligence LaboratoryHeadquarters

2018Est.

Focus Areas

MCP Tools

Human-in-the-Loop Training

Browser Automation

Connect

AI Tools by Web Machine Learning Community Group

(1)

WebMCP

Web API for AI Tool Registration

MCP Tools HITL Training Browser Automation

Discussions

No discussions yet

Be the first to start a discussion about Web Machine Learning Community Group

Latest News

11/10/2025

Products & Services

Web Neural Network API (WebNN)

2018 (Incubation start)

A low-level Web API for machine learning inference in the browser, providing hardware acceleration for neural networks.

Model Loader API

Incubating

An API to enable efficient loading and management of machine learning models in web environments.

Task-specific APIs

Ongoing

Incubation of higher-level APIs for common ML tasks like Face Detection, Barcode Detection, and Image Classification.

Market Position

Standardizing cross-platform Web APIs for ML inference to compete with fragmented native platform-specific libraries.

Leadership

Founders

Anssi Kostiainen

Intel Corporation; long-time W3C contributor and staff member; proposed the WebML CG in 2018.

Rijubrata Bhaumik

Intel Corporation; initial supporter and key contributor to Web Neural Network API.

Zoltan Kis

Intel Corporation; initial supporter and contributor.

Executive Team

Anssi Kostiainen

Chair

Intel Corporation; lead editor of multiple W3C specifications.

Dominique Hazaël-Massieux

W3C Staff Contact

W3C staff member facilitating the group's activities and coordination.

Board of Directors

Google LLC

Participating Organization

Intel Corporation

Participating Organization

Microsoft Corporation

Participating Organization

Founding Story

Proposed in October 2018 in response to feedback from a TPAC breakout session to provide a low-level standard for ML hardware acceleration in browsers.

Business Model

Revenue

N/A (Community Group)

Revenue Model

Non-profit consortium group; activities are community-driven and free to join/participate in. W3C is funded by member organization dues.

Pricing Tiers

Standard Participation

Community Group participation is open to all without charge.

N/A (W3C is a non-profit)

Target Markets

Industries & Segments

Web developers
Browser vendors
Hardware manufacturers
ML framework authors

Use Cases

Image recognition
Natural language processing
On-device privacy-preserving ML
Agentic web experiences

Notable Customers

TensorFlow.js
ONNX Runtime Web
Hugging Face

Quick Facts

Headquarters

MIT Computer Science and Artificial Intelligence Laboratory

Founded

2018

Entity Type

W3C Community Group

Total Funding

N/A (Funded via W3C membership dues)

Investors

Intel, Google

Office Locations

Cambridge

Sophia Antipolis

Tokyo

Beijing

History & Milestones

2025-09-25

Group re-chartered to focus on next-generation features including agentic web and built-in models.

2025-11

Face-to-face meetings at TPAC 2025 in Kobe, Japan, focusing on AI agents and browser integration.

2021-04-27

W3C Launches the Web Machine Learning Working Group to standardize graduated APIs like WebNN.

2020-06

W3C Workshop on Web & Machine Learning held to explore standardization needs.

2018-10-03

Web Machine Learning Community Group proposed and launched by Anssi Kostiainen.

Key Capabilities

Hardware-accelerated inference

Neural network primitives

Support for custom models

Integration with JS frameworks

Integrations & Partnerships

Platform Integrations

Chrome
Edge
Firefox
Safari (WebKit)

Key Partnerships

W3C WebAssembly Working Group

W3C GPU for the Web Working Group

Connect

Website

w3.org/community/webmachinelearning/

GitHub

webmachinelearning

AI Topics

Web Machine Learning Community Group focuses on these topics:

MCP Tools(1)

Human-in-the-Loop Training(1)

Browser Automation(1)

Back to all developers

Web Machine Learning Community Group

To make Machine Learning a first-class web citizen by incubating and developing a dedicated low-level Web API for machine learning inference in the browser.

Visit Website

At a Glance

MIT Computer Science and Artificial Intelligence LaboratoryHeadquarters

2018Est.

Focus Areas

MCP Tools

Human-in-the-Loop Training

Browser Automation

Connect

AI Tools by Web Machine Learning Community Group

(1)

WebMCP

Web API for AI Tool Registration

MCP Tools HITL Training Browser Automation

Discussions

No discussions yet

Be the first to start a discussion about Web Machine Learning Community Group

Latest News

11/10/2025

WebML WG/CG F2F Agenda - TPAC 2025 (Kobe, Japan)

github.com

04/27/2021

W3C Launches the Web Machine Learning Working Group

w3.org

10/03/2018

Call for Participation in Machine Learning for the Web Community Group

w3.org

Products & Services

Web Neural Network API (WebNN)

2018 (Incubation start)

A low-level Web API for machine learning inference in the browser, providing hardware acceleration for neural networks.

Model Loader API

Incubating

An API to enable efficient loading and management of machine learning models in web environments.

Task-specific APIs

Ongoing

Incubation of higher-level APIs for common ML tasks like Face Detection, Barcode Detection, and Image Classification.

Market Position

Standardizing cross-platform Web APIs for ML inference to compete with fragmented native platform-specific libraries.

Leadership

Founders

Anssi Kostiainen

Intel Corporation; long-time W3C contributor and staff member; proposed the WebML CG in 2018.

Rijubrata Bhaumik

Intel Corporation; initial supporter and key contributor to Web Neural Network API.

Zoltan Kis

Intel Corporation; initial supporter and contributor.

Executive Team

Anssi Kostiainen

Chair

Intel Corporation; lead editor of multiple W3C specifications.

Dominique Hazaël-Massieux

W3C Staff Contact

W3C staff member facilitating the group's activities and coordination.

Board of Directors

Google LLC

Participating Organization

Intel Corporation

Participating Organization

Microsoft Corporation

Participating Organization

Founding Story

Proposed in October 2018 in response to feedback from a TPAC breakout session to provide a low-level standard for ML hardware acceleration in browsers.

Business Model

Revenue

N/A (Community Group)

Revenue Model

Non-profit consortium group; activities are community-driven and free to join/participate in. W3C is funded by member organization dues.

Pricing Tiers

Standard Participation

Community Group participation is open to all without charge.

N/A (W3C is a non-profit)

Target Markets

Industries & Segments

Web developers
Browser vendors
Hardware manufacturers
ML framework authors

Use Cases

Image recognition
Natural language processing
On-device privacy-preserving ML
Agentic web experiences

Notable Customers

TensorFlow.js
ONNX Runtime Web
Hugging Face

Quick Facts

Headquarters

MIT Computer Science and Artificial Intelligence Laboratory

Founded

2018

Entity Type

W3C Community Group

Total Funding

N/A (Funded via W3C membership dues)

Investors

Intel, Google

Office Locations

Cambridge

Sophia Antipolis

Tokyo

Beijing

History & Milestones

2025-09-25

Group re-chartered to focus on next-generation features including agentic web and built-in models.

2025-11

Face-to-face meetings at TPAC 2025 in Kobe, Japan, focusing on AI agents and browser integration.

2021-04-27

W3C Launches the Web Machine Learning Working Group to standardize graduated APIs like WebNN.

2020-06

W3C Workshop on Web & Machine Learning held to explore standardization needs.

2018-10-03

Web Machine Learning Community Group proposed and launched by Anssi Kostiainen.

Key Capabilities

Hardware-accelerated inference

Neural network primitives

Support for custom models

Integration with JS frameworks

Integrations & Partnerships

Platform Integrations

Chrome
Edge
Firefox
Safari (WebKit)

Key Partnerships

W3C WebAssembly Working Group

W3C GPU for the Web Working Group

Connect

Website

w3.org/community/webmachinelearning/

GitHub

webmachinelearning

AI Topics

Web Machine Learning Community Group focuses on these topics:

MCP Tools(1)

Human-in-the-Loop Training(1)

Browser Automation(1)

Back to all developers