Releases·Tool Release
browser-llm-fit: Check if an AI model fits the browser
A developer released browser-llm-fit, a JavaScript library that probes a client's WebGPU hardware limits — such as maxStorageBufferBindingSize and shader-f16 support — and ranks which AI models can run in the browser before weights are downloaded. It targets crashes from oversized or unsupported model loads in WebGPU tabs.
1 engagement2 sourcesThu, Sep 3, 2026, 04:59 PM
AI Overview
- browser-llm-fit probes client hardware limits — maxStorageBufferBindingSize, shader-f16 — and ranks browser-executable models before download; fit('SmolLM2-135M') returns fits and '45-65 tokens/sec'
- Available live, on GitHub (hemanth/browser-llm-fit), and via npm; author requests feedback on odd GPU setups and mobile WebGPU
Related
Releases·Tool Release
Software Mansion releases runntime, a TypeGPU-based framework for running AI models on the user's GPU in WebGPU environments
Software Mansion released ruNNtime, a framework for running AI models on the user's own GPU via WebGPU, built on TypeGPU. Kernels are written in TypeGPU and compiled to WGSL at runtime, so no WASM binary is needed and the same code runs in the browser, React Native, and Electron. It offers a 'zoo' of one-function-per-task models for speech, vision, and NLP, positioning itself as an alternative to transformers.js for on-device web inference.
Software Mansion released ruNNtime, a framework for running AI models on the user's own GPU via WebGPU, built on TypeGPU. Kernels are written in TypeGPU and compiled to WGSL at runtime, so no WASM binary is needed and the same code runs in the browser, React Native, and Electron. It offers a 'zoo' of one-function-per-task models for speech, vision, and NLP, positioning itself as an alternative to transformers.js for on-device web inference.
37 engagement1 sourcegithubMon, Sep 14, 2026, 01:28 PM
Products·launched
LLMSpeed — estimate local LLM tokens/s before buying a GPU
A small web tool that estimates how fast a local LLM will run (tokens per second) on hardware before you purchase the GPU. It targets people considering local LLM setups who want a feel for inference speed without buying hardware first. The author notes results are estimates and will vary per user setup.
A small web tool that estimates how fast a local LLM will run (tokens per second) on hardware before you purchase the GPU. It targets people considering local LLM setups who want a feel for inference speed without buying hardware first. The author notes results are estimates and will vary per user setup.
3 engagement1 sourceredditSat, Aug 29, 2026, 06:17 PM
Products·launched
Pure WASM 4-bit LLM inference engine for browser, no GPU or server
A WebAssembly-based inference engine that runs a 350M-parameter LiquidAI LFM 2.5 model entirely in the browser with 4-bit quantization. It requires no GPU, no server, and no network at inference time, enabling private on-device AI chat for users without powerful hardware.
A WebAssembly-based inference engine that runs a 350M-parameter LiquidAI LFM 2.5 model entirely in the browser with 4-bit quantization. It requires no GPU, no server, and no network at inference time, enabling private on-device AI chat for users without powerful hardware.
4 engagement1 sourceredditWed, Jul 15, 2026, 02:28 AM
Products·launched
1-bit LLM inference in browser via WebGPU
A WebGPU-accelerated demo that runs a 1-bit quantized LLM entirely in the browser. It enables local, private text generation without server calls, targeting developers and researchers exploring edge inference.
A WebGPU-accelerated demo that runs a 1-bit quantized LLM entirely in the browser. It enables local, private text generation without server calls, targeting developers and researchers exploring edge inference.
22 engagement1 sourcehackernewsThu, Jul 16, 2026, 04:50 PM
Products·launched
RigMatch: benchmarks Ollama models on your own hardware and ranks them
RigMatch is a local tool that reads your GPU, VRAM, RAM and disk to determine which Ollama models fit your machine, then runs selected models through an identical question set and ranks them. It scores speed and answer quality and shows full transcripts. It helps people running local LLMs stop guessing whether a model will perform well on their hardware.
RigMatch is a local tool that reads your GPU, VRAM, RAM and disk to determine which Ollama models fit your machine, then runs selected models through an identical question set and ranks them. It scores speed and answer quality and shows full transcripts. It helps people running local LLMs stop guessing whether a model will perform well on their hardware.
1 engagement1 sourceredditTue, Sep 8, 2026, 05:17 PM