ONNX Runtime Web
PulseAugur coverage of ONNX Runtime Web — every cluster mentioning ONNX Runtime Web across labs, papers, and developer communities, ranked by signal.
5 day(s) with sentiment data
-
Local LLMs now run in browsers via WebGPU, enhancing privacy
The WebGPU API, now stable across major browsers like Chrome, Edge, and Firefox, enables the execution of local Large Language Models (LLMs) directly within the browser. This advancement, coupled with optimizations in O…
-
Browser-based horror app Localghost uses local LLMs for personalized ghost stories
Localghost is a novel horror storytelling application that runs entirely in the browser, designed to create personalized ghost stories. Users can select a scare level, provide personal photos and notes about their envir…
-
AI agent apps run entirely in-browser using Laya decision model
A developer has created four AI agent applications that run entirely within a web browser using Laya, a 421M-parameter decision model from ConvAI Innovations, and WebGPU. These applications require no server or API keys…
-
Developer embeds LLM in website to run locally in browser
A developer has successfully integrated a Large Language Model (LLM) directly into their personal website, allowing it to run within the visitor's browser without requiring a backend server or API keys. This approach en…
-
LLMs run in-browser via WebGPU, cutting inference costs
A new approach allows large language models to run directly within a web browser using WebGPU and Transformers.js, eliminating the need for server-side inference. This method, demonstrated in the BEINGS chatroom applica…
-
AI text and image detectors compressed for client-side browser use
A developer has successfully compressed an AI text detector, originally based on the RoBERTa model, from approximately 500 MB to 97 MB, enabling it to run client-side in a web browser. Additionally, an image detector wa…
-
Developer creates local Chrome extension to filter YouTube feed
A developer has created a Chrome extension that utilizes a small, locally running model to filter YouTube content directly within the browser. This extension, which does not require a server or API key, allows users to …
-
Guide Explains Running AI Models Directly in Web Browsers
A practical guide has been released detailing how to run AI models directly within a web browser using the Transformers.js library. This guide explores the capabilities of Transformers.js, which allows for the execution…
-
AI models now run directly in browsers using ONNX Runtime Web
ONNX Runtime Web is enabling complex AI tasks like background removal and feature extraction to be performed directly within a web browser. This client-side processing eliminates the need for powerful backend servers, r…
-
WebGPU unlocks browser GPU power for massive parallel computing
WebGPU is a new web API that allows developers to leverage the massive parallel processing power of graphics processing units (GPUs) directly from the browser. Unlike its predecessor, WebGL, which was limited to renderi…
-
AI image models like Stable Diffusion now run in-browser via TypeScript
New advancements in ONNX Runtime Web and WebGPU are enabling complex AI models like Stable Diffusion and Flux to run directly within web browsers using TypeScript. This shift moves AI model execution from dedicated back…
-
Ternary Quantization Shrinks Super-Resolution Transformer to 668 KB for Browser Use
Researchers have successfully applied BitNet-style ternary quantization to a super-resolution transformer model, resulting in a significantly smaller model size. The quantized model, which uses weights of -1, 0, or +1, …
-
Moebius image inpainting model ported to browser using Claude Code
Simon Willison successfully ported the Moebius 0.2B image inpainting model to run in a web browser using Claude Code. The process involved converting the model to ONNX format and leveraging WebGPU for browser-based exec…
-
ProxyFace adds local, emotional avatars to AI chats
ProxyFace is an open-source project that adds a local, expressive avatar to AI interactions. It utilizes a small, on-device emotion model and eye-tracking to make the avatar react to AI output and the user's gaze. The p…