Browser-local AI

Semantic Playground

Explore how embeddings recognize meaning beyond matching keywords. One local model powers three practical experiments.

Semantic Playground

Local processing

Model not loaded

First use needs internet and about 25 MB. Loading may reuse cached files.

Try:

First text

61/1500

Second text

Your text stays in this browser. The model is downloaded from Hugging Face and may be cached locally; the model host receives a normal file request but never the text entered here. No comparison is saved by Best-AI.org.

Learn how local AI works

Understand model downloads, quantization, privacy boundaries, and the limits of running AI on your own hardware.

Open the Local AI course

Model transparency

Model: Xenova/all-MiniLM-L6-v2
Revision: 751bff37182d3f1213fa05d7196b954e230abad9
Hosting: Fetched once from Hugging Face (pinned revision), then runs on-device from the browser cache
Download: ~25 MB on first use
Runtime: WASM in a Web Worker
License: Apache-2.0

View model source