browser-llm on HackerShow

browser llm demo

SHOW HN2025

WebGPU enables local LLM in the browser – demo site with AI chat

Shown by andreinwald on Aug 2, 2025. Click the card to turn it over.

145points

54comments

#514of 3,837 in 2025

Top 14%of 2025 by points

How HN took it

MixedNo clear lean

4 positive, 13 neutral or asking, 0 mixed, 11 critical, out of 28 comments about the project. Thread score -0.19.

Performance: 1 for, 6 againstExecution: 0 for, 4 againstIdea: 3 for, 0 against

Each row is a top-level comment and its replies, in HN's order. Select a cell to open that comment on HN.

positive neutral or mixed critical about another commenter or something else the maker

Answers from the small model are poor quality
From 3 comments
- Fun demo but the model that's used seems to be pretty stupid: > What's the best way to get to space? >> Unfortunately, it's not currently possible for humans to travel to space…
  echoangle on HN
- The answer is still terrible for the model size. Maybe it's the 4 bit quantization, smaller models tend to react worse to that For reference, [1] is what stock quen3-0.6B would…
  wongarsu on HN
- I asked "Why is the sky blue?" and got back a response of "coppia RR TalentDevExpressincer+'.//////////////////////////////////////////////////////////////////////// cha…
  AndrewDucker on HN
Not working in Safari, which lacks WebGPU support
From 2 comments
- Its not working. Loading model: waiting WebGPU error: WebGPU is NOT supported on this browser. Im on macOS and Safari.
  imatrix on HN
- Same on iOS and mobile Safari - not working.
  imatrix on HN
Model download is about 500 MB, costly on mobile data
From 1 comment
- Not everyone enjoys unlimited data plans, and 500 MB is schon a lot.
  pjmlp on HN
Integrated GPU is used, not the discrete one
From 1 comment
- Any way to make this use my actual GPU instead of the terrible integrated one?
  201984 on HN

Load local GGUF model files instead of downloading to browser cache
From 1 comment
- Does anyone know why all of these WebGPU LLM demos have you download the models to browser storage rather than letting you open a gguf already on your local drive? I have several…
  apitman on HN
Keep the text input always visible on screen
From 1 comment
- very cool. improvement would be if the input text box is always on screen, rather than having to manually scroll down as the screen fills.
  andsoitis on HN

Is this just a wrapper around web-llm?
From 2 comments
- Looks like this is a wrapper around: https://github.com/mlc-ai/web-llm Which has a full web demo: https://chat.webllm.ai/
  maxmcd on HN
- Is this correct? It doesn't seem so to me, either from the way it works or from what little of the code I've looked at... But I don't have time to do more than the quick glance I…
  swores on HN
Which model does the demo use?
From 1 comment
- Which model does the demo use?
  cgdl on HN

MLC web-llm demo chat
From 1 comment
- Looks like this is a wrapper around: https://github.com/mlc-ai/web-llm Which has a full web demo: https://chat.webllm.ai/
  maxmcd on HN
Ollama
From 2 comments
- What's the performance of a model like vs an OpenAI API? What's the comparable here? Edit: I see it's same models locally that you'd run using Ollama or something else. So…
  asim on HN
- Yes, its very similar to Ollama app, and Llama-3.2-1B model used
  andreinwald on HN
Chrome Prompt API built-in Gemini Nano
From 1 comment
- There is a Prompt API in development that's available in both Chrome and Edge to give access to a local LLM. Chrome extensions have access to it and I believe websites can…
  scottfr on HN
Hugging Face transformers.js
From 1 comment
- Very cool! For some awesome examples of running different types of models in webgpu check out the examples folder here. https://github.com/huggingface/transformers.js
  dataviz1000 on HN