transformers-js

Run state-of-the-art ML models in JavaScript across browsers and Node.js.

Updated Aug 23, 2026

Search Tags:#javascript #pipeline #nodejs #browser #inference #ml #transformers.js

Ready to run

npx skills add https://github.com/domattioli/DomI --skill transformers-js-domattioli

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: transformers-js
Source: https://github.com/domattioli/DomI/tree/main/plugins/huggingface-skills/skills/transformers-js
Command: npx skills add https://github.com/domattioli/DomI --skill transformers-js-domattioli

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Running state-of-the-art ML models typically requires a Python environment or servers. Transformers.js lets developers execute NLP, computer vision, audio, and multimodal models directly in JavaScript, enabling client-side and server-side inference without Python dependencies.

Core Features & Use Cases

Cross-runtime ML: run models in browsers, Node.js, Bun, or Deno with WebGPU or WASM backends.
Pipeline-driven inference: load tasks such as text-classification, image-classification, translation, and more via a simple pipeline API.
Model discovery, quantization, caching, and per-component options to balance performance and memory.
Examples include sentiment analysis in a web app, image classification in a mobile browser, or offline inference in a Node.js service.

Quick Start

Install the package and load a pipeline for a given task and model, then run inference.

Frequently Asked Questions about transformers-js

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I run ML models in JavaScript without a Python backend?▼

Run ML models in JavaScript without Python backends by using Transformers.js to execute NLP, computer vision, audio, and multimodal tasks directly in browsers or Node.js via a pipeline API.

Can I use WebGPU or WASM for inference in a web browser?▼

WebGPU and WASM backends are supported for running inference directly in web browsers, allowing flexible execution of state-of-the-art models without server-side dependencies.

How do I load a text-classification or image-classification pipeline in Node.js?▼

Load a pipeline in Node.js by installing the package and specifying a task like text-classification or image-classification with a model, then run inference to process inputs.

What is the best way to manage caching and quantization for ML inference in JavaScript?▼

Manage caching and quantization for ML inference in JavaScript by configuring environment settings, which allow per-component options, loading modes, and resource management to balance performance and memory.

Does Transformers.js work across different JavaScript runtimes like Bun and Deno?▼

Transformers.js works across multiple JavaScript runtimes including Bun and Deno, enabling cross-runtime ML inference with support for both WebGPU and WASM backends.

Why use a JavaScript pipeline API instead of a Python server for ML inference?▼

A JavaScript pipeline API enables client-side and server-side ML inference without Python dependencies, supporting offline execution in Node.js services and direct browser deployment.