A working review of LingoChunk, the audio-to-flashcards and shadowing tool from Show HN. What it nails, where its scope stops, and who it actually fits.
LingoChunk is a clever, focused tool for one job: turning native-language audio into study material. Word-level timestamps let it mine vocabulary by base form, make audible cloze flashcards you can export to Anki, and loop any passage for shadowing — a genuinely nice workflow for serious learners. It is an early, solo-built project (15 input languages, experimental Japanese/Chinese, prompts still in progress) and it is purely a study tool: it generates and publishes no content. Judge it as a learning aid, where it earns its place, not as a content tool.
Most "turn audio into flashcards" tools are either a thin transcription wrapper or a paywalled app that hides the good part. LingoChunk, shared as a Show HN on 25 June 2026, is neither — it is a solo developer's working study tool, built first for their own German and Greek, then opened up. The reason it is worth a real review is that it does a few specific things unusually well.
The foundation is transcription with word-level timestamps. From that, LingoChunk mines vocabulary grouped by base form (lemma) with real spoken examples, generates cloze flashcards where the target word is blurred on the front and "beeped" in the example audio, holds them in a built-in spaced-repetition system, and exports to Anki. Click the first and last word of any passage and it loops that segment with an adjustable gap and repeat count for shadowing. Select a line and an AI helper returns a quick grammar or nuance note. That is a thoughtful chain for anyone who learns from native audio.
This review scores LingoChunk as what it is: a language-study tool. I will not mark it down for failing to be a video editor or a social scheduler, because it never claimed to be either. But I will be clear about the boundary, because the search results lump AI tools together and it is easy to assume an "audio-to-flashcards" tool might also make content for a language channel. It does not. The honest read: LingoChunk is a strong study aid that stops at the flashcard and the loop.
Everything here reflects LingoChunk's documented scope on its Show HN post and site as of the review date. It is an early, fast-moving project, so treat the specifics as a snapshot.
LingoChunk is a browser-based tool that converts spoken audio into language-study material. Its core is transcription with word-level timestamps: it transcribes the audio, extracts the words, finds their base forms (lemmas), and groups example sentences by lemma, so you study vocabulary in dictionary form with real spoken context. From that it builds cloze-style flashcards — the target word blurred on the front, revealed on the back, optionally "beeped" in the example audio so you test recall against sound — in a built-in spaced-repetition system, with Anki export. A shadowing mode loops a passage you select with a configurable gap and repeat count for pronunciation drilling, and the developer also uses it as a plain pronunciation player. An AI helper sends a selected line to a predefined "explain grammar" or "explain nuances of meaning" prompt. It was built by a solo developer (Hacker News user "alder") and supports 15 input languages — with Japanese and Chinese as recent, experimental additions — and more than 30 output languages for explanations; the transcription back-end covers many more languages than were tested. It generates and publishes no content: there is no video or image generation, no caption engine, no carousels, no scheduler. Pricing was not publicly detailed at launch beyond a free try page seeded with public-domain audio.
The clearest fit is a serious self-directed learner who studies from native audio — podcasts, audiobooks, language-course recordings — and wants vocabulary, flashcards, and pronunciation practice from the same source. Anki users who want cloze cards mined automatically (rather than hand-cropping clips in Audacity and pasting translations, as one commenter described doing before) will get the most out of it, as will anyone who values shadowing as a speaking drill. It is a weaker fit for absolute beginners who need a structured curriculum, and it is the wrong tool entirely for a language creator whose goal is producing and publishing lessons — there is nothing in it for that job.
| Dimension | Score | Why |
|---|---|---|
| Audio → flashcards (cloze, lemma grouping) | 4.3 / 5 | The standout. Word-level timestamps and base-form grouping turn raw audio into well-organized, audible cloze cards. |
| Shadowing / pronunciation drilling | 4.2 / 5 | Select a passage and loop it with adjustable gap and reps — a clean, focused way to drill speaking from native audio. |
| Anki export & SRS | 4.0 / 5 | Built-in spaced repetition plus Anki export slots into trusted study workflows; the audio "beep" is not yet in the export. |
| AI grammar / nuance explanations | 3.5 / 5 | Handy in-context notes on a selected line, but the developer describes the prompts as still experimental. |
| Language coverage | 3.5 / 5 | 15 input and 30-plus output languages, with Japanese and Chinese still experimental; the back-end can cover more over time. |
| Transcription accuracy / timestamps | 4.0 / 5 | Word-level timing is reliable enough to anchor loops and cloze cards; edge cases noted in experimental languages (e.g. pinyin). |
| Maturity & polish | 3.0 / 5 | Early, solo-built, fast-moving. Useful today, with rough edges (tiny UI on large displays, in-progress features). |
| Pricing / value | 3.5 / 5 | Free try page; full pricing not publicly detailed at launch, so value is hard to score precisely yet. |
| Content production & publishing | 1.0 / 5 | Out of scope by design. No video, images, captions, carousels, or social publishing of any kind. |
LingoChunk launched with a free try page seeded with public-domain audio (LibriVox, Wikimedia, and FSI course recordings) and did not publish detailed pricing in its Show HN post. That makes a precise value verdict premature, so this review scores it on capability rather than on a price-to-feature ratio that does not yet exist publicly. Treat the pricing rating as provisional and check the site for current terms.
What I can say fairly is that the workflow it automates has real value to the right learner. Several commenters described doing this by hand before — cropping audio clips in Audacity, pulling translations from textbook PDFs, and assembling Anki cards manually. If LingoChunk replaces an hour of that per study session, a reasonable price would be easy to justify for a committed learner. The open question is simply where the developer lands the paid tier, and how much of the workflow stays in the free try.
The honest framing: as an early, single-developer project, LingoChunk is being shaped in public, and its pricing is part of what is still moving. That is normal for a Show HN tool and not a strike against it — just a reason to verify the current terms yourself before assuming a number.
| Use case | Fit | Why |
|---|---|---|
| Self-study from podcasts and audiobooks | Strong | Mining vocabulary, making cloze cards, and shadowing from native audio is exactly what it is built for. |
| Anki users who want auto-generated cloze cards | Strong | Automates the manual crop-and-paste card-making many learners do by hand, with timestamps and lemma grouping. |
| Pronunciation / shadowing practice | Strong | Loop-a-passage mode with adjustable gap and reps is a clean, purpose-built speaking drill. |
| Studying experimental languages (Japanese, Chinese) | OK | Supported but newer — the developer flags them as experimental, with edge cases like pinyin still being refined. |
| Absolute beginners needing a curriculum | Weak | It is a study workbench over audio you choose, not a structured beginner course with a learning path. |
| Producing language lessons for an audience | Weak | No generation or publishing — it makes private study material, not postable content. |
| Multi-platform content distribution | Weak | There is no scheduler or publisher; distribution is entirely out of scope. |
LingoChunk and Kompozy are not competitors — they sit on opposite ends of a language creator's day. LingoChunk is the study desk: where you mine audio, drill cards, and shadow a native speaker to get the language into your own head. Kompozy is the publishing end: where a phrase you just learned becomes a finished lesson an audience sees.
The interesting overlap is recurring output. A creator's weekly study routine in LingoChunk naturally surfaces teachable moments — a confusing grammar point, five real uses of one word, a phrase worth shadowing on camera. Hand that theme to Kompozy and it generates the lesson set (a Persona Short modeling the pronunciation with captions, a carousel for the rule, a thread, a "phrase of the week" newsletter, a blog for search) in one consistent teaching voice via a Persona Brief, then schedules it across nine platforms. Put a topic pool of lesson themes on autopilot and a private study habit becomes a public series you barely touch.
The honest boundary: if you only want to learn, LingoChunk alone is the answer and Kompozy is not relevant. If your bottleneck is producing and posting lessons at the volume a channel needs, no study tool closes that gap — that is the specific job Kompozy was built for.
For a self-directed learner who studies from native audio, yes — it automates vocabulary mining, cloze flashcards, and shadowing from one source, and exports to Anki. It is an early, solo-built project with experimental edges, so test it on the free try page first, and it is not the tool if you want a structured beginner curriculum.
LingoChunk is a browser tool that turns native-language audio into study material: it transcribes with word-level timestamps, mines vocabulary grouped by base form, makes cloze flashcards (Anki-exportable), loops passages for shadowing, and offers AI grammar/nuance notes. It was shared as a Show HN in June 2026.
It transcribes the audio, finds the base form (lemma) of each word, groups example sentences by lemma, and generates cloze cards where the target word is blurred on the front and revealed on the back — and "beeped" in the example audio. Cards live in a built-in spaced-repetition system and export to Anki.
You click the first and last word of a passage in the transcript and LingoChunk loops that audio segment with a configurable gap and repeat count, so you repeat after the native speaker to drill pronunciation and rhythm.
At launch, 15 input languages (with Japanese and Chinese as experimental additions) and more than 30 output languages for explanations. The transcription back-end supports many more languages than were tested, so coverage may grow — check the site for the current list.
It has a free try page seeded with public-domain audio so you can test it. As an early, solo-built project it did not publish detailed pricing at launch, so check lingochunk.com for the current terms.
No. LingoChunk makes private study material — flashcards, loops, and grammar notes — and does not generate video, images, or social posts or publish anywhere. To turn what you study into published lessons across platforms, you would use a content engine like Kompozy.