webgpu

Vercel has open-sourced vgpu.sh, an agent-native development tool for WebGPU that enables shader execution in browsers and headless Node.js, with features for CI testing and agent integration.

Spline V2 is a completely rebuilt 3D editor platform for the agentic era, introducing AI Agent Mode, WebGPU engine, and custom scripting features.

Galaxium is a tool for exploring the universe using WebGPU technology, offering interactive and graphical experiences.

Pngine is a declarative format and runtime for WebGPU that uses S-expressions to simplify declaring and validating WebGPU configurations, enabling cross-platform sharing and export to formats like HTML or PNG with embedded runtime.

The author rebuilt Mozilla's discontinued Orbit AI browser summarizer as a privacy-focused local tool called Apogee, which connects to Ollama for inference and supports summarization of articles, videos, and forums.

Marc demonstrates depth-aware light injection in VisionCamera using DINOv2 and Apple Vision framework, enabling real-time rendering with react-native-webgpu.

S1-mini is a 600M-parameter LLM that cleans up speech-to-text transcripts by removing errors and adding punctuation, running locally in the browser via WebGPU and Transformers.js.

The article introduces motion as a new creative medium for shaders, using WebGPU compute shaders to extract motion through frame differencing and create visual effects like trails and motion blur.

HashAgent lets you package an AI agent into a shareable URL that runs locally in the browser using WebGPU, enabling privacy-preserving, serverless AI agent distribution.

jax-js is an open-source ML library and compiler for the web, bringing JAX-like numerical computing to JavaScript with WebGPU and WebAssembly kernels, all running client-side.

parakeet.wgsl enables fast, accurate NVIDIA Parakeet TDT 0.6B V2 speech transcription entirely in the browser using raw WebGPU compute shaders and SIMD WASM, with a live demo and open-source library.

A cyberpunk FPS tech demo that runs entirely in the browser, built in 3 days via vibe coding with multiple LLMs. It features procedural generation, WebGPU rendering, and fully synthesized audio.

The author details how they built an in-browser poker solver by using LLMs to generate custom WebGPU kernels instead of relying on a general tensor library, achieving over 10x speedup and demonstrating a paradigm shift where cheap generation can replace library abstraction.

Academa Studio brings 3Blue1Brown's Manim animation engine to the browser using WebGPU, enabling AI-native Manim editing with no local setup.

A free interactive book teaching graphics programming with WebGPU in JavaScript, covering from basics to advanced topics like GPU compute and Gaussian splatting.

The author explains how Web3DSurvey.com collects real-world WebGL/WebGPU capability data for under $3/month using a cost-efficient architecture with BigQuery and GET requests to avoid CORS preflight.

A 1-bit LLM (Bonsai) is now runnable in the browser via WebGPU, enabling efficient on-device inference.

Bonsai 27B is a 1-bit dense large language model that can run locally in a browser using custom WebGPU kernels, enabling efficient on-device inference.

Bonsai 27B achieves 93% size reduction via 1-bit quantization while retaining 90% intelligence, enabling local browser inference with custom WebGPU kernels.

Google announces LiteRT.js, a high-performance JavaScript runtime for running AI models directly in the browser using WebAssembly and hardware acceleration, as an evolution from TensorFlow.js.