Semantic Search
Free tools › Productivity › Semantic search
Search by meaning, not by exact words
Type what you want in your own words and the directory finds the page that answers it, even if the words are different. Everything runs in your browser: an embedding model turns your question into a vector and compares it with the index.
- Your query never leaves the tab
- Model cached after the first search
- 16 pages indexed
Semantic search over the directory
Results will appear here, ranked by similarity, not by keyword.
The first search downloads the model once (about 23 MB) and indexes these pages in your browser. After that it answers offline in under a second.
Keyword search vs semantic search
The same question, two very different behaviours.
Keyword
Looks for the exact letters. If you write «cut out», a page titled «remove background» never appears.
Semantic
Compares meaning. «Cut out a product photo» lands on the background remover, with other tools behind it.
Private
Both the index and the query stay in the tab. No search engine sees what you are looking for.
Frequently asked questions
- What is it searching?
- The pages of this directory: the free tools and the main guides. The list lives in the page itself, so it grows as new pages are added.
- Why does the first search take longer?
- Because the embedding model (about 23 MB) is downloaded once and then every page is turned into a vector in your browser. After that, each search takes well under a second.
- Does it send my query anywhere?
- No. There is no server and no search API: the model runs locally and the comparison happens in memory, in your tab.
- Does it work offline?
- Once the model and the index are in the cache, yes. Reload the page and search without a connection.
- Can I use it for my own site?
- Yes. It is the same approach: a list of pages, the model and a cosine similarity. The code is plain JavaScript with no build step.
More free tools
The rest of the lab, no sign-up and also inside your browser.