Plaud Importer

⚠️ Early Alpha

This plugin is in early alpha and under active development. Expect bugs, UI changes between releases, and occasional rough edges. It talks to an undocumented, reverse-engineered Plaud web API that Plaud can change at any time.

Before installing, please read What to know before you install. It explains the main tradeoffs in plain language, and exactly what the plugin will and will not touch in your vault.

No warranty. Use at your own risk, and you may want to test in a non-critical vault first.

Please report issues

I use this plugin myself every day, so I often catch breakage quickly, but I am one person on one account, and Plaud can change their API for some accounts or regions before it affects mine. If something breaks for you, an issue report is the single biggest help, and may be the first signal I get.

1. File a GitHub issue at ckelsoe/obsidian-plaud-importer/issues.
2. Include your Obsidian version, plugin version, OS, and steps to reproduce.
3. Attach a debug log. See Capturing a debug log below. Auth headers are redacted automatically.
4. Attach a screenshot or the generated note if the output looks wrong.

Feature requests and API-shape observations are welcome too. Open an issue or discussion.

Import meeting recordings, AI summaries, transcripts, and attachments from Plaud.AI into your Obsidian vault as markdown notes.

Each recording becomes a single note with frontmatter metadata, a Plaud-generated summary, and the transcript with chapter navigation. You choose whether the transcript sits under a heading, in a collapsed callout, or in its own file. Images, mind-maps, and other Plaud artifacts land in a matching -assets folder next to the note.

What to know before you install

Plaud does not offer an official API yet, so this plugin works by using Plaud's own (undocumented) web service the same way the Plaud website does. That is what makes it possible at all, and it brings a few tradeoffs worth understanding up front. None of them put your existing notes at risk.

The tradeoffs

• It can break when Plaud changes their website or API. There is no official, supported API, so Plaud can change how their service works at any time and an import may suddenly stop working. Why it works this way: it is the only way to get your recordings into Obsidian today. What protects you: the plugin shows a clear error instead of writing broken notes, and when Plaud ships an official API the plugin will move to it.
• Your sign-in token is like a key to your whole Plaud account. The plugin stores the session token your login uses. It is not a limited, read-only key. Why: that is the only credential Plaud's web service provides. What protects you: it is kept in Obsidian's secure secret storage, never written to your settings file, never logged, and never sent anywhere except Plaud.
• You sign in through Plaud's own website. Clicking Sign in opens Plaud's real site in a separate window so you can log in normally; Google and Apple logins use your normal browser instead (see Connect your Plaud account). Why: you never hand your password to the plugin. What protects you: the plugin reads only the token your logged-in session already uses; your password is typed into Plaud's own page, not the plugin.
• It is an unofficial tool, not affiliated with Plaud. What protects you: it reads only your own data, and only when you ask it to. Following Plaud's terms of service is your responsibility.
• Desktop only. Sign-in and import rely on desktop features that Obsidian Mobile does not provide.

What it will not do

• It only adds notes and attachments inside the output folder you choose, and never edits or deletes anything outside that folder. (To spot already-imported recordings it does read the frontmatter of notes inside your output folder. See Permissions and access.)
• It replaces an already-imported note only when you choose Overwrite (or confirm it under "Ask each time").
• It does not rewrite Plaud's transcript or summary text. It renders Plaud's own data into a note.
• The only change it can make in your Plaud account is a recording's title, and only if you turn on the optional title write-back and either confirm the prompt or opt into automatic updates (see Renaming recordings). It never edits or deletes anything else in Plaud. With that setting off, the plugin is read-only against Plaud.

What it does

• One-click sign-in. Connect your Plaud account from settings without copying a token from the browser console. A sign-in window opens, you log in to Plaud normally, and your session token is captured automatically. Your password is never seen by the plugin. (Manual token paste is still available as a fallback.)
• Test connection. A button in settings makes one lightweight call to Plaud and tells you whether your token works, so you can confirm you are signed in (or learn you need to sign in again) without running a full import.
• Lists your recent Plaud recordings in a modal with scroll-to-load pagination.
• Lets you pick which to import via checkboxes, single or multi-select.
• Per-recording artifact selection. Before a multi-import you can tick/untick transcript, summary, attachments, mindmap, card, and screenshots independently.
• Writes one markdown note per recording with:
  • YAML frontmatter (Plaud ID, date, start and end time, duration, speakers, tags, Plaud web URL), with times written in each recording's own capture time zone, plus a layered set of optional fields surfaced from Plaud's flat GPT-5 schema (plaud-headline, plaud-category, plaud-industry, plaud-language, plaud-template, plaud-model, plaud-note-id, plaud-summary-id, plaud-summary-version), emitted only when present, never load-bearing.
  • Plaud's AI summary, plus any additional summaries the recording has (such as a newer beta summary on the new portal), each in its own section
  • An AI Suggestions section pulled from Plaud's ai_suggestion field when the response includes one (separate from the main summary)
  • A Template outputs section that renders any extra AI templates you generated in Plaud (Key Points, Daily Journal, Meeting Summary, and so on), one foldable subsection per template, instead of dropping them as unreadable .bin attachments
  • A Screenshots section with any photos you captured during the recording (new portal), each labelled with its time into the recording and downloaded into the note's assets folder
  • The transcript, laid out the way you pick in settings (see Transcript rendering). The default is a heading-based section with a chapter index that jumps to each chapter, and Back to Chapters links

Requirements

• Obsidian 1.11.4 or newer. Required for the SecretStorage / SecretComponent APIs used to handle the Plaud token securely.
• Desktop only (isDesktopOnly: true). The authentication path depends on Electron APIs (including the separate sign-in window) that are not available on Obsidian Mobile. This restriction will be lifted when Plaud ships a public OAuth API (see Plaud API status below).
• A Plaud.AI account with access to the recordings you want to import.

Installation

From Community plugins (recommended)

Open the Plaud Importer listing and click Install, or inside Obsidian:

1. Open Settings → Community plugins → Browse.
2. Search for Plaud Importer and click Install.
3. Click Enable.

Via BRAT (beta and pre-release versions)

BRAT installs and auto-updates plugins straight from GitHub. It is handy before this plugin lands in the catalog, or for tracking pre-release builds.

1. Install Obsidian42 - BRAT from the Obsidian community plugins catalog and enable it.
2. Open the command palette and run BRAT: Add a beta plugin for testing.
3. Paste the repository URL: https://github.com/ckelsoe/obsidian-plaud-importer.
4. BRAT downloads the latest release (main.js, manifest.json, styles.css) and installs Plaud Importer.
5. Enable Plaud Importer in Settings → Community plugins.
6. To update later: BRAT: Check for updates to all beta plugins, or set BRAT to auto-check on Obsidian start.

Manual install

1. Download main.js, manifest.json, and styles.css from the latest release.
2. Copy all three files into <your vault>/.obsidian/plugins/plaud-importer/.
3. Reload Obsidian (or disable/re-enable the plugin) and enable Plaud Importer in Community plugins.

Configuration

Open Settings → Community plugins → Plaud Importer and configure:

Connect your Plaud account

The plugin reads your recordings using your Plaud web-session token. You never give the plugin your Plaud password. You sign in to Plaud directly, and the plugin captures only the session token your logged-in session already uses.

Because Plaud has no official API, which method works depends on how you log in to Plaud. Pick the one that matches your account. Both store your token the same way; the difference is only how the token is captured.

If you use the plugin in more than one vault: each vault keeps its own separate Plaud sign-in, so you sign in once per vault. This is deliberate. Vaults no longer interrupt each other's sessions, Clear sign-in in one vault leaves the others signed in, and you can point different vaults at different Plaud accounts if you want to.

Sign in with email

Use this if you log in to Plaud with an email address and password.

1. In Settings → Community plugins → Plaud Importer, under Sign in, click Sign in next to Sign in with email.
2. A separate sign-in window opens with Plaud's website. Log in with your email and password.
3. Once you reach your library, the plugin captures your token automatically and the window closes. The Plaud token status changes to "connected."

Google and Apple logins do not complete in this window. If your account uses Google or Apple, see the next method, and the tip there about accounts that also have a password.

Sign in with Google or Apple

Use this if you log in to Plaud with single sign-on (SSO) through a Google or Apple account. Those logins only work in a real web browser, so this method signs you in there and hands the token back to Obsidian. There is a quick one-time setup (saving a bookmark), then a few clicks each time.

If your Plaud account also has a password, use email sign-in instead. This applies when you created your Plaud account with an email and password and linked Google or Apple to it later. Log in with that email and password through Sign in with email above. That method renews itself in the background for about 30 days between sign-ins. A Google or Apple session captured with the bookmark below is less reliable: on Plaud 4.0 the plugin tries to renew it in the background but Plaud can still end it early, and on Plaud 3.0 the plugin cannot renew it at all and it usually lasts about 24 hours.

If you created your Plaud account with Google or Apple, use the bookmark method below. Plaud support has confirmed that a password cannot be added to an account registered through Google or Apple, so email sign-in is not available for it.

One-time setup (do this once):

1. In the plugin settings under Sign in, find Sign in with Google or Apple and click Set up bookmark. A small page opens in your default web browser.
2. That page shows a button labeled Plaud → Obsidian (v2). Drag that button up onto your browser's bookmarks bar. If you do not see a bookmarks bar, press Ctrl+Shift+B (Windows/Linux) or Cmd+Shift+B (Mac) to show it, then drag the button onto it.
3. The bookmark is now saved. You will not need to repeat this step.

If you use more than one vault: the bookmark is built for the vault you set it up in, and the setup page names it. Obsidian delivers this kind of link to the vault it is addressed to, so it works even when another vault window has focus. To use the plugin in a second vault, run Set up bookmark there too and keep one bookmark per vault.

Coming from an earlier version? Re-do this one-time setup and replace your old Plaud → Obsidian bookmark. The new one sends the token to Obsidian for you instead of asking you to copy and paste it, and it finds the token on accounts where the old one found nothing.

Each time you connect (whenever your session expires):

1. Back in the plugin settings, click Launch sign-in to capture token. A short reminder pops up; read it and click Open my browser now. Plaud opens in your browser.
2. Sign in to Plaud with Google or Apple if you are not already signed in.
3. Click the Plaud → Obsidian bookmark you saved. Your browser asks whether to open Obsidian; allow it. Obsidian saves the token and the Plaud token status changes to "connected."

If your browser holds more than one saved Plaud sign-in, the plugin tries them against Plaud and keeps the one that actually works, so you do not have to know which is which.

If Obsidian does not open in step 3, the bookmark falls back to showing a line of text in a pop-up box. Select the whole line, copy it (Ctrl+C or Cmd+C), switch back to Obsidian, and click Paste token from clipboard. That line carries the same sign-ins the bookmark would have sent, so the plugin still picks the one that works.

If nothing at all happens when you click the bookmark, make sure you are signed in to Plaud in that tab first, then click the bookmark again.

Once the Plaud token status shows "connected", close that Plaud browser tab, and do not use Plaud's Log Out. Logging out ends the same session Obsidian is holding, so it would sign you out here too.

Starting over: the Clear sign-in button clears the plugin's stored token and its own embedded sign-in session, so you can connect a different account or recover from a stuck state. It does not sign you out of Plaud in your normal web browser.

How long a sign-in lasts depends on the sign-in method. A session from the email sign-in window renews itself in the background for about 30 days, then asks you to sign in again. A session captured from a Google or Apple login is less reliable: on Plaud 4.0 the plugin tries to renew it in the background but Plaud can still end it early, and on Plaud 3.0 the plugin cannot renew it and it usually lasts about 24 hours. If your account also has a password, the email sign-in is the steadier choice (see the tip above). The plugin settings show your session's actual expiry in the status line under Plaud token, and when the plugin can tell your session is short, it says so at sign-in. When imports stop working, sign in again with the method for your account. If a session lapses during background auto-sync, the pause notice's Reconnect button reopens the right sign-in for your account type, so an SSO account is not sent to the email window where Google and Apple logins do not work.

The token is stored in Obsidian's per-vault secret storage either way. It is never written to data.json and does not travel through Obsidian Sync. Switching vaults requires re-connecting.

Plaud 3.0 and Plaud 4.0: Plaud is moving accounts from Plaud 3.0 to Plaud 4.0, and both work. The plugin detects which one your account is on each time you sign in, and switches to 4.0 by itself when Plaud moves your account. The Plaud version setting under Sign in shows what was detected. Leave it on Auto unless support asks you to change it.

Regional accounts (EU and others) need no extra setup. If Plaud routes your account to a regional server, the plugin detects it on the first import and remembers it.

Test connection

After signing in, click Test connection to confirm your token reaches Plaud. It makes one lightweight call and reports either that your token works or the specific problem (for example, an expired token that needs a fresh sign-in). Use it any time imports start failing to tell quickly whether you need to sign in again.

Output folder

Folder inside your vault where imported notes are written. Defaults to Plaud. Nested paths work (Archive/Plaud/2026).

Subfolder template

By default every note lands directly in the output folder. To keep a growing library organized, set a Subfolder template that files each note into a subfolder built from the recording's date. Leave it empty to keep the flat layout.

Text inside {{ }} is a date format written in Moment style, the same syntax the core Daily Notes plugin uses. Text outside the braces is kept as-is. Any forward slash (/) in the result starts a new nested folder level, whether you type it outside the braces or a date format produces it inside them, so {{YYYY}}/{{MM}} and {{YYYY/MM}} both make a year folder with month folders inside it. Dashes and spaces are fine inside the braces (they are part of the date format); keep your own words (plain letters) outside them, because letters inside are read as date tokens.

Common tokens (case matters; the same set works in the note-name field):

Token	Expands to
{{YYYY}}	Year, for example 2026
{{MM}}	Month, 01 to 12
{{MMMM}}	Month name, for example June
{{DD}}	Day, 01 to 31
{{dddd}}	Weekday name, for example Monday
{{WW}}	ISO week number, 01 to 53
{{Q}}	Quarter, 1 to 4

Examples (folder for a June 4 2026 recording):

Template	Resulting folder
{{YYYY-MM}}	2026-06 (one folder)
{{YYYY}}/{{MM}}	2026/06 (a 2026 folder containing a 06 folder)
{{DD}}-{{MM}}-{{YYYY}}	04-06-2026 (one folder, day-first order)
{{YYYY}}/W{{WW}}	2026/W23 (a 2026 folder containing a W23 folder)
{{YYYY}}/{{MM MMMM}}	2026/06 June (two tokens in one {{ }})

You choose the order and the separators, so any date layout works, and a whole layout can live in one {{ }}. A live preview under the field shows the resulting folder as you type, and insert-token buttons add the correct token at the cursor. The folder is built from the recording's own date, so re-importing always resolves to the same place. A recording with no usable date goes to an _undated subfolder.

The template applies to new imports; notes you already imported stay where they are until a re-import overwrites them. Each note's -assets folder follows it into the same subfolder. If you change the template later, a re-import that overwrites an existing note moves it, with its -assets folder, into the new subfolder instead of creating a duplicate.

Note name

By default each note is named YYYY-MM-DD <title>, using the recording's own date. To control the format, set a Note name template that builds the name from the same {{ }} Moment date formats as the subfolder setting, plus a {{title}} token for the recording title. Put the date before or after the title, in any order and with your own separators:

Template	Resulting note name
{{YYYY}}-{{MM}}-{{DD}} {{title}}	2026-07-03 Team sync
{{title}} {{YYYY}}-{{MM}}-{{DD}}	Team sync 2026-07-03
{{MMM D, YYYY}} - {{title}}	Jul 3, 2026 - Team sync (one combined date token)
{{title}}	Team sync (no date in the name)

Preset buttons fill in ISO, US, and EU orders, insert-token buttons add a token at the cursor, and a live preview shows the resulting name as you type, or type your own. Keep your own words outside the braces, since bare letters inside are read as date tokens. The recording's own date replaces any date already in the Plaud title, so a title like 04/13 Meeting becomes the recording's date, and a title with no date gets one. A note name has to be a valid filename, so a template is rejected if its result would contain a character a file name cannot hold (a slash, colon, square bracket, asterisk, question mark, angle bracket, pipe, or double quote), be a reserved device name such as CON, start or end with a dot or space, or run over 200 characters. A # or ^ in the name (usually from a recording title) is replaced with the replacement character instead, because Obsidian cannot link to a note or its attachments under a name that contains them. The date property inside each note stays YYYY-MM-DD regardless, for Dataview and sorting.

Like the subfolder setting, this applies to new imports and leaves notes you already imported alone until they are re-imported. When a re-import overwrites a note whose name no longer matches (because the recording's title changed in Plaud, or you changed this template, the subfolder template, or the replacement character, or the old name contained # or ^), the note moves to its new name together with its -assets folder. A note the re-import skips keeps its name. To rename a note by hand, see Renaming recordings.

Renaming recordings

You can rename an imported recording from Obsidian and keep everything in sync:

• Run the command Plaud Importer: Rename recording on the note, or right-click the note and choose Rename imported recording. Enter a new name, and the plugin renames both the note and its -assets folder so embedded images keep working.
• If you rename a Plaud note directly in the file explorer, the plugin moves its -assets folder to match automatically.
• When you rename a recording in Plaud itself, its note and assets folder are renamed on the next sync (or manual re-import), instead of leaving a stale-named note.

Update the recording title on rename (off by default) controls whether a rename is also pushed back to Plaud:

• Off: the Rename recording command asks each time whether to also update the title in Plaud, and a file-explorer rename stays entirely local.
• On: both the command and a file-explorer rename update the title in Plaud automatically, with no prompt.

The title sent to Plaud is your new note name exactly, so Plaud matches what you see in Obsidian. If your Plaud session has expired, the plugin offers to sign you in and then finishes the update, so a rename is not lost. This is the only change the plugin writes back to Plaud.

Duplicate handling

What the importer does when a note for a recording already exists in the output folder:

• Skip. Leave the existing note untouched. The importer reports it as skipped in the summary.
• Overwrite. Replace the imported section of the note and clear its matching -assets folder before re-downloading. Text you wrote outside the imported section is kept (see Your own text in an imported note). A confirmation modal fires once per import batch before any overwrite happens.
• Ask each time (default). Prompt per duplicate at write time. The modal shows the recording title and the exact target path, warns that the note's imported section AND its -assets folder will be replaced, and offers Overwrite, Skip, or (in multi-item batches) Overwrite all remaining, Skip all remaining, and Cancel import.

Your own text in an imported note

Everything the plugin writes below the frontmatter sits between two hidden marker lines:

<!-- plaud:managed:start -->
# 2026-04-14 Morning standup
...summary, attachments, transcript...
<!-- plaud:managed:end -->

Reading view and Live Preview hide the markers. In Live Preview a marker line shows again while your cursor or selection is on it, so you can still see or delete it; source mode always shows them. When a recording is re-imported, only the text between them is replaced. Anything you write above the start marker or below the end marker stays exactly as you left it, and so do frontmatter properties you added (see Preserve unknown frontmatter on re-import in settings). Edits inside the markers are replaced by Plaud's current version.

• A note imported before this feature has no markers. Its next re-import rebuilds it whole and adds them, so protection starts after that re-import.
• If a marker is deleted, duplicated, or moved out of order, the plugin cannot tell its text from yours, so it does not update that note. A manual import reports it in the import summary, and background auto-sync skips it while the recording keeps its Update available badge. Restore one start and one end marker, or delete both to let the next import rebuild the whole note.
• If you have written text below the end marker, a re-import does not collapse the transcript heading, because the fold would hide your text too. The callout layout is unaffected.
• To detect that a recording changed (for example from a script or another plugin), watch the plaud-version-ms property. It holds Plaud's version stamp for the recording, so it changes when Plaud's copy of the recording changes.

Default artifact selection

What the "Review artifacts first" checklist starts with when you begin a multi-import: transcript, summary, attachments, mindmap, card, and audio. Uncheck artifacts you never want to pull by default; you can always override per-batch. Audio is off by default because the files are large (roughly 15 MB per hour of recording); turn it on only if you want the original audio saved and embedded in each note.

Auto-sync

Optional background sync, off by default. When enabled, the plugin checks Plaud on a schedule and imports new recordings on its own using your default import options, with no import window open.

• Enable auto-sync. Turns the background job on or off.
• Sync interval. How often it checks, from every 15 minutes to once a day.

It also re-imports recordings you changed in Plaud since importing them (an edited transcript, corrected speaker names, a regenerated summary), including older recordings you edited today. A re-import overwrites the imported section of that note and its downloaded artifacts with Plaud's current version, so edits you made inside that section are replaced. Text you wrote above or below it is kept. Only recordings that actually changed are touched; unchanged notes are never modified. This tradeoff is called out in the settings.

If your Plaud session expires, auto-sync pauses with a single notice and resumes automatically when you reconnect (Test connection, a manual import, re-entering your token, or toggling auto-sync off and on). The notice's Reconnect button opens the sign-in that matches your account: the email window for password logins, or the browser bookmarklet flow for Google and Apple accounts.

Backfill for existing notes: notes imported before this feature shipped do not carry the version marker auto-sync compares against, so their Plaud-side edits are not detected until the marker exists. Run the command Plaud Importer: Backfill version markers for auto-sync once after enabling auto-sync to make your existing library edit-detectable. It only adds a frontmatter marker and never rewrites note content.

Recording sources

Optional filter, off by default, that limits import to the sources you choose. Useful when one of your Plaud devices records things you do not want in your vault, so background auto-sync stops pulling them in.

• Only import from selected sources. The master switch. Off means every recording is imported (the original behavior).
• Sources. Click Load devices to fetch the devices paired to your Plaud account, then uncheck any whose recordings you want to skip. A separate App, desktop and imported recordings checkbox covers recordings made in the Plaud app or imported, which have no device.

The filter applies to both manual import and auto-sync. It uses blocklist behavior: a source is imported unless you uncheck it, so a device you pair later is included by default until you change it here.

Seeing and saving the source. Whether or not the filter is on, the import list shows a small source chip on each row (the device name, or "App" for a recording made in the Plaud app or imported) so you can tell which device a recording came from while choosing what to import. To keep that on the note itself, add a {{device}} token to an Extra frontmatter row in settings, for example device: {{device}}, and each imported note records the device it came from for filtering and Dataview.

Tags and keywords

• Tag mode. Chooses which sources land in each note's tags: frontmatter: no tags, your custom tags only, Plaud tags (the default), or all (which also adds Plaud's AI keywords as plaud/... tags).
• Custom tags. Comma-separated tags added to every imported note. Defaults to plaud-meeting.
• Keep AI keywords as note property. When AI keywords are kept out of tags:, optionally write them to a keywords: property instead. Off by default: Plaud's keyword list can run to hundreds of low-value entries per recording, which buries the tags that matter and adds noise to every note. Turn it on if you want the full list.

Transcript rendering

• Transcript layout. Where each note's transcript goes:

  • In the note, under a heading (default). A Transcript heading with a chapter index that jumps to each chapter heading.
  • In the note, in a collapsed callout. Keeps the transcript out of the note's heading outline. Chapters show as a plain list and bold titles, with no jump links, because Obsidian cannot jump to a chapter inside a collapsed callout.
  • Separate file, embedded in the note. The transcript goes in Transcript.md in the note's -assets folder and is embedded under the note's Transcript heading, so it still reads inline. In the file, chapters are real headings with working links.
  • Separate file, linked from the note. Same file, but the note only links to it, so the note stays short.

  The layout applies to notes you import or re-import after you change it. Existing notes keep their layout until they are re-imported. The plugin rewrites Transcript.md on every re-import, so edit the note rather than that file. If you switch back to an in-note layout, re-importing the note moves its old Transcript.md to the trash. A note whose name contains # or ^ keeps its transcript in the note under a heading, because Obsidian cannot link to a file whose path contains those characters.

• Fold transcript by default. Imported notes open with the transcript heading collapsed so the summary is what you see first. The callout layout is always collapsed, so this does not apply to it.

• Transcript heading level. Sets which H-level the Transcript heading uses (in the heading layout, chapters render one level deeper). Not used by the callout layout.

Unprocessed and trashed recordings

• Write placeholder for unprocessed recordings (on by default). When Plaud has a recording but no transcript or summary yet, write a small placeholder note (recording ID plus a link back to Plaud) instead of recording a bare failure. A later successful import replaces the placeholder automatically, even under Skip or Ask each time. Turn it off to leave such recordings as plain failures with no file written.
• Show trashed recordings (off by default). The import list hides recordings that are in your Plaud trash, matching the Plaud app. Turn this on to bring them back when you want to import something you trashed in Plaud but still want in your vault. The same control is mirrored in the import dialog's filter bar as Hide trashed, and any trashed recording that is shown carries a Trashed badge so it is easy to tell apart from your live recordings. Accounts on Plaud 4.0 do not have this setting or the Hide trashed toggle: Plaud's 4.0 recording list contains only active recordings, so there is no trash to show.

Import dialog

• Filter bar. Four toggles at the top of the dialog, all phrased as "Hide" so a ticked box always means hidden: Hide already-processed (on by default; hides recordings already imported and unchanged), Hide updates available (off by default; hides imported recordings that changed in Plaud, which stay shown by default because re-importing them is real work), Hide ignored (on by default), and Hide trashed (on by default; not shown on Plaud 4.0 accounts). Your choices are remembered between sessions.
• Ignore a recording. Each row has an eye button. Ignoring a recording drops it from the list (while Hide ignored is on) and stops background auto-sync from ever importing it, and it works on recordings you never imported. Turn off Hide ignored to see ignored recordings again and un-ignore any of them.
• Empty list. Once auto-sync has imported your newest recordings, opening the dialog with Hide already-processed on can leave nothing to show. The dialog pages past the hidden recordings on its own to find any still worth importing, and when there is genuinely nothing left it says so and points you to the filters, instead of showing a blank list.
• Auto-close summary (on by default). The post-import summary closes itself after a short countdown. Only a fully successful run auto-closes; any failure keeps the window open so the error list stays visible, and clicking inside the window cancels the countdown.
• Auto-close delay. Seconds to wait before the summary closes (default 20).

Appearance

• Show ribbon icon (on by default). Toggle the left-rail icon that opens the import window. Turn it off to launch imports only from the command palette. The change takes effect immediately.
• Ribbon icon. Pick the icon from a curated set (audio-lines, mic, headphones, podcast, and others). A live preview sits next to the dropdown and the ribbon updates immediately.

Debug logging

Off by default. When on, captures API request/response metadata and parsed results into an in-memory buffer for troubleshooting. Auth headers are stripped. Payloads may contain transcript text, speaker names, and recording metadata, so only enable when you are preparing a bug report. Use the command Plaud Importer: Debug: copy debug log to clipboard to export the session.

Using it

1. Click the audio-lines ribbon icon on the left rail, or run the command Plaud Importer: Import recent recordings.
2. The list shows recordings still worth importing; recordings you already imported are hidden by default. Use the Hide filter bar at the top to change what is shown, and each row's eye button to ignore a recording you never want.
3. Scroll the recording list to load older pages (handled automatically as you scroll).
4. Tick the recordings you want.
5. Click Import N recordings (or Review artifacts first to uncheck specific artifacts for this batch).
6. Watch the per-file progress counter. A final Notice summarizes how many were imported, skipped, or failed; failures are listed in the modal with a Copy button for bug reports.

Capturing a debug log

When something breaks, a debug log is the most useful thing to attach to an issue. It records the API requests and responses the plugin made (auth headers are stripped automatically; payloads may contain transcript text and recording metadata, so review before posting).

1. Settings → Community plugins → Plaud Importer, turn on Debug logging.
2. Reproduce the problem (run the import that failed, etc.).
3. Open the command palette and run Plaud Importer: Debug: copy debug log to clipboard.
4. Paste it into your GitHub issue. Turn Debug logging back off when you are done.

Plaud API status

⚠️ This plugin currently uses Plaud's undocumented web API under api.plaud.ai. The endpoints in use are GET /file/simple/web for listing recordings, POST /ai/transsumm/{id} for transcript and legacy summary retrieval, and GET /file/detail/{id} for the richer detail bundle (polish-revision summaries, AI keywords, mindmap and card attachments, and AI suggestions). The only write endpoint is PATCH /file/{id} to update a recording's title, used solely by the optional title write-back described in Renaming recordings. None of these are officially published or supported by Plaud.

The listing and /ai/transsumm/{id} endpoints were informed by prior community work, most directly rsteckler/applaud, with cross-validation against JamesStuder/Plaud_API. Those projects established that the data was reachable from a logged-in web session and gave a useful starting point for the auth and listing flow. The richer /file/detail/{id} surface used by this plugin (the transaction_polish polish-revision summary, auto_sum_note, ai_suggestion, the GPT-5 frontmatter schema fields, and the pre-signed S3 URLs for mindmap, card, and attachment assets), along with the encoding quirks (millisecond timestamps on the listing endpoint, the empty-JSON POST body shape on /ai/transsumm/{id}), were verified and extended through direct inspection of live Plaud responses during this plugin's development.

What this means for you:

• The plugin may break without warning if Plaud changes URL shapes, response schemas, or authentication. When it breaks, the import modal will surface a clear error (auth failure, parse error, HTTP 4xx/5xx) rather than silently corrupt notes, but some releases may need to wait for a plugin update.
• Your Plaud token is a full web-session JWT, not a scoped API key. Treat it with the same care as your Plaud password. The plugin stores it in Obsidian's SecretStorage, never in data.json, and never logs it.
• When Plaud ships a public OAuth API, this plugin will switch to that surface and deprecate the reverse-engineered path. A separate OAuthPlaudClient stub is already stubbed in the codebase for the migration. The mobile-only restriction (isDesktopOnly) exists because of the current auth path and will be re-evaluated at that point.

I am actively monitoring Plaud's developer announcements and waitlist for the official API. If you hear anything, please open an issue.

Permissions and access

Plaud Importer is desktop-only, runs on your device, and has no telemetry or maintainer server. It talks only to Plaud, and only when you ask it to. Obsidian's plugin scan discloses a few capabilities. Here is exactly what each is and why it exists:

• Network access to Plaud. The plugin calls Plaud's web API (api.plaud.ai, or your regional host) to list and fetch your recordings, summaries, transcripts, and attachments, and downloads attachment files from the CDN hosts Plaud's responses point at. It contacts no other third party, and only when you trigger an import, scroll to load more recordings, or download attachments. The one write it can make to Plaud is updating a recording's title, and only when you enable the optional title write-back and confirm or opt into automatic updates (see Renaming recordings); with that setting off, every call to Plaud is read-only.
• Sign-in window. When you click Sign in, the plugin opens Plaud's own website (web.plaud.ai) in a separate window so you can log in normally. For Google or Apple logins, you sign in through your normal web browser instead and hand the token back with a bookmarklet. Either way, your password is entered into Plaud's page and is never seen by the plugin; it reads only the session tokens your logged-in session already holds, checks them against Plaud to find the one that works, and stores that one via SecretStorage. The sign-in runs in a private session isolated from Obsidian's other web views.
• Vault file enumeration. To show the "Imported" badge and avoid duplicate imports, the plugin lists your vault's note paths and reads the frontmatter of notes inside your output folder to find ones it previously created (tagged with a plaud-id). It does not read the contents of unrelated notes.
• Clipboard. The plugin writes to your clipboard for the Copy buttons (copy debug log, copy an error or failure list for a bug report). It reads your clipboard in exactly one place: when you click Paste token from clipboard during sign-in. That read happens only on that click, is used only to look for a Plaud sign-in token, and is never stored or sent anywhere if it is not one. Nothing reads your clipboard in the background.
• Secret storage. Your Plaud token is stored via Obsidian's SecretStorage (per-vault, not synced); data.json holds only a reference id, never the token.

All vault writes go through Obsidian's Vault API (Vault.create, Vault.process) with no direct filesystem access, and nothing here sends your data anywhere except Plaud.

See PRIVACY.md for the full privacy policy and liability disclaimer, and SECURITY.md for the security policy. The plugin is provided "AS IS" with no warranty (see LICENSE); it is not affiliated with or endorsed by Plaud.AI, and you use it at your own risk.

Troubleshooting

• "No Plaud token configured". Re-check the Plaud token field in settings. If your token expired, connect your Plaud account again (click Sign in).
• "Plaud rejected your token". Your web session likely expired or you signed out of Plaud. Click Sign in again to refresh the token (or re-paste it manually), then use Test connection to confirm it works.
• "Could not reach Plaud.AI". Network or DNS issue on your side, or Plaud is down. Retry from the modal's Retry button.
• "Plaud returned data in an unexpected shape". Plaud changed their API. File an issue with the debug log attached (see Debug logging).
• Import silently "skipped". Your duplicate handling was set to Skip and the note already existed. Switch to Ask each time (default since 0.2.0) or Overwrite.
• A brand-new recording will not import yet. A recording does not become available to import until you have opened Plaud (web or app) and its cloud has finished finalizing the recording. Until then the plugin cannot see it. Open Plaud, let it finish processing, then run the import again. This is a Plaud-side step, not a plugin limitation.

Community

Questions, ideas, and general discussion happen on Discord. For anything that needs tracking, a GitHub issue is still the better home.

Support and issues

Please report bugs and feature requests at github.com/ckelsoe/obsidian-plaud-importer/issues. Include:

• Your Obsidian version
• Your plugin version (shown in Community plugins)
• The debug log if reproducible (strip anything sensitive)

Contributing

See CONTRIBUTING.md for local build and test instructions.

License

MIT. See LICENSE.

Acknowledgments

Inspired by prior community work on Plaud's web API. rsteckler/applaud and JamesStuder/Plaud_API demonstrated that recordings, transcripts, and summaries are reachable from a logged-in web session, and gave this plugin a head start on the listing and transcript-retrieval flow. The richer detail-bundle surface (polish-revision summaries, AI keywords, mindmap and attachment asset URLs, AI suggestions, GPT-5 schema fields) and the response-shape and timing quirks were worked out by inspecting live Plaud responses during this plugin's development.

Thanks to Obsidian for the SecretStorage and SecretComponent APIs that make storing the Plaud token securely a non-issue.