Summarize Saved HTML

Extract readable text from saved HTML, select original sentences, and download a reading copy with the complete extracted text and native report.

Processing stays in your browser; code and assets download as needed. Original and checkpoint downloads retain source content, including omitted values. Reset and Exit stop this session’s owned work; they do not establish secure erasure or anonymity.

This flow needs JavaScript. If it does not load, refresh the page and try again.

About this flow

Extract readable text from saved HTML, select original sentences, and download a reading copy with the complete extracted text and native report.

Limitations

Choose saved HTML as UTF-8 TXT or paste it (500,000 UTF-16 units, 2 MB). No page or linked resource is fetched. The original parser omits scripts, styles, templates, noscript and comments. Malformed HTML may be repaired; it does not identify main article content or preserve images and layout. The summarizer accepts extracted text up to 100,000 units and 1–500 punctuation-separated sentences. Select 1–50 original sentences. Ranking uses English-oriented ASCII content words and .!? boundaries, then restores source order. This does not paraphrase or verify facts. The final ZIP retains the complete extracted text, native report and selected sentences. Download original HTML separately in this workflow. Private visible text is not redacted. If extraction is empty, too long or has too many sentences, the complete extracted TXT remains downloadable.
ExistingHTMLparser is bounded synchronous work and cannot be interrupted mid-call. Originalsummary30secondworker/100000units/400000combinedUTF8bytes/4MiBreport unchanged. Owned90secondconnection and8MiB/4memberZIP.