IGTranscript

Instagram transcript summarizer: why the summary is a second pass

There is no summariser here, and the reason is structural rather than commercial. The thing that turns speech into text in this tab is a speech recogniser, and 22,349 bytes of app.js contain 0 occurrences of chunk_callback, 0 of max_new_tokens, 0 of temperature and 0 of the word “summar”: there is no second model in the page and no setting that would make the first one do a second job. What the page hands you is the whole thing — measured on 27 October 2026, a 600.3-second clip came back as 1,194 words, 111 segments and 8,173 characters on screen, 6.85 characters for every word spoken, with nothing dropped. A summary is a different operation on top of that: it deletes words without replacing them, so it belongs after the transcript, in the hands of something that was built to write.

InVideo WW

The rest of this page is the evidence: which lines turn audio into text and why none of them can compress, the arithmetic in characters and bytes that a summary has to work against, what a real summariser would add to a download that is already 43.6 MB, what the two of the three sites that sell this feature say their output actually is, and four ways to get a summary from a transcript you already hold.

What the transcript pass can do, and what it has no dial for

The recognition call is one line and it has four keys. pipe(pcm, { language, task: 'transcribe', return_timestamps: true, chunk_length_s: 30 }) at line 278 of a 579-line file — a language, a task pinned to transcribe, timestamps on, and a 30-second window. That is the entire surface. A max-token limit, a sampling temperature, a beam width, a no-repeat penalty: all four are absent, so there is no knob that would make the model say less.

Two passes side by side: the transcript pass is one model, four keys and no deletion; the summary pass is a second model that removes words it cannot put back. One pass, then a different pass The recogniser writes; the summariser deletes. They are not the same operation with a different setting. Pass 1 — recognition whisper-base q8, wasm, 76,894,612 bytes audio → 16,000 Hz mono 4 keys: language, task, stamps, 30 s every word, 1,194 of them nothing deleted, nothing added then Pass 2 — summarising a model this page does not load reads the text, writes other text keeps some, drops most fewer words than were spoken and the dropped ones are gone for good A summariser is a second reader. It cannot be the same model with one more flag, because no flag exists.
Where the two passes sit. The left box is the 22,349 bytes of app.js measured on 12 November 2026: one recogniser, four keys on the call at line 278, no token limit, no temperature, no chunk_callback. The right box is what a summary needs, and it is absent from the file altogether — 0 occurrences of the word “summar” in 22,349 bytes. The arrow is the only place an outside tool can join in, and it carries the same 1,194 words and 8,173 characters the measured 600.3-second run produced.

Why a summary has to delete, and what deleting costs in characters

Speech recognition is close to lossless; summarising is not, by definition. The measured run is the yardstick: 1,194 words over 600.3 seconds is 119.3 words a minute, and 8,173 characters of stamped text is 6.85 characters per word. A summary that shrinks that by any meaningful factor has to remove words, and once a word is removed from a transcript no later pass can tell you what it was.

What a summariser would add to a download that is already 43.6 MB

The two checkpoints the page offers are 76,894,612 bytes for the better-accuracy model (79.7 MB as the page quotes it) and 43,649,619 bytes for the smaller one (43.6 MB), and both are already int8-quantised and running on WASM rather than on a graphics card. A summariser is a language model with its own weights, which on the same machine would arrive as a third download, not as an option inside the first.

What the two sites that sell this feature say it outputs

Three sites dominate this category, and two of them publish a summariser page for Instagram; the third does not have one in the benchmark set held here. Both pages agree on the shape of the output and disagree about the name of one of its parts, and neither one puts a number on how much shorter the summary is than the transcript.

Four ways to summarise a transcript you already hold

None of these needs a summariser inside the transcription tool, and each one keeps the full transcript as the fallback if the summary turns out to be the wrong shape for what you want.

The question behind the query

Someone searching for an Instagram transcript summarizer is usually not looking for a feature list; they are looking at a Reel that is too long to read. The honest answer to that is the one this page can measure: the transcript of a ten-minute Reel is 8,173 characters and arrives in about 0.42 seconds of processing for every second of audio, and the summary is a separate, lossy step that belongs wherever you already write. The tool that hears and the tool that shortens are two tools, and keeping them apart is what lets you reread the first one when the second one gets it wrong.

Last checked: 12 November 2026. Measured on this machine (Windows 11, one AMD EPYC 9754 core with 2 threads and 3.9 GB of RAM) unless marked otherwise: the 600.3-second run — 1,194 words, 111 segments, 8,173 characters on screen, 251.5 seconds of recognition inside 271.9 seconds end to end, 180 MB of JavaScript heap — is the run logged on 27 October 2026, and the 39,999-character render test and the buildSrt()/buildVtt() byte counts come from a sandbox run of the deployed app.js on 12 November 2026 (22,349 bytes, 579 lines, md5 0431b8cb5a4e008607f720f78766258e). Everything else is arithmetic on constants in that file — RATE = 16000 at line 236, task: 'transcribe' at line 280, chunk_length_s: 30 at line 282 — or on the model’s own generation_config.json, read on the same day: 1,194 ÷ 600.3 × 60 = 119.3 words a minute; 8,173 ÷ 1,194 = 6.85 characters a word; 600.3 × 16,000 × 4 = 38,419,200 bytes of model input, which is 0.4996 of the 76,894,612-byte checkpoint; 1,194 × 0.25 = 298.5 words. The checkpoint sizes are the repository’s own file listing for onnx-community/whisper-base and onnx-community/whisper-tiny. The competitor sentences are reproduced exactly as they appear in this site’s benchmark directory, which held 30 content pages totalling 227,427 characters of body text on 12 November 2026, with each page’s heading list stripped before counting. The 25 per cent compression and the 200-words-a-minute reading speed are illustrations on top of measured figures, not measurements of any product, and no search volume, difficulty or ranking estimate is given here because none was retrieved.

If what you want is the transcript before the summary, the accuracy page counts the wrong words in a measured run, the timestamps page explains the 15-character prefix that makes a summary jumpable, the SRT and VTT page covers the cue boundaries you can split on, and the repurposing page takes the transcription further into a draft than a summary does.