SpeechdashHelp Center

Reader overview

Every control on the document reader at /library/[documentId], including billing and how each tool works.

The document reader is the full-screen page at /library/[documentId]. The main app sidebar is hidden. This is where you listen, edit, ask questions, create versions, translate, share, and export.

This page maps every area of that screen. Linked guides go deeper on a single tool.

speechdash.com/library/doc-1

Introduction

This sentence is highlighted during playback.

The following paragraph continues…

Reader: header, side toolbar, sentence highlight, playback bar.

Layout

AreaWhat it does
HeaderBack to home, clickable title (rename), version switcher
Side toolbar (left)New version, outline, edit, export, More (⋯)
Main columnMarkdown reader, edit mode, PDF processing, or a failed-import panel
Playback barPlay, skip sentences, timeline, speed, voice (Eco/Cloud), mic, Ask questions
Ask questions panel (right)Opens from the playback bar; resizable
Dialogs and sheetsRename, new version, translate, share, download audio, unsaved changes

A first-visit tour highlights the toolbar, player, Eco/Cloud voice, and progress slider. Dismissing it once stores sd-document-tour-seen in the browser; it does not show again on that device.

ControlBehavior
BackReturns to home. Playback stops. If you have unsaved edits, Speechdash asks to save or discard
TitleOpens rename (title, version name, document language)
Version switcherLists every version. A spinner means import or generation is still running. Switching pauses playback and loads that version’s text, voice settings, and last sentence

Each version keeps its own text, skip list, voice settings, and resume position. See Switch versions and Resume playback.

If the document id is unknown or you cannot access it, the reader shows Document not found and a link home.

Side toolbar

The floating left toolbar stays visible in read and edit mode. Full item list: Document side toolbar.

ControlWhat happensCredits
New versionOpens the version sheet: Natural, Lecture, Conversational, or Original (copy / language-model translation)Tokens + 1 credit surcharge per generation. Estimate shown before confirm
OutlineHeadings H1–H3. Disabled when there are none. Click a heading to jump playbackFree
Edit / Stop editingToggles edit mode. Playback pauses when you enter editFree to edit. Enhance on a selection uses tokens + 1 credit
ExportPDF, DOCX, TXT always; SRT when timings exist; Download audio (MP3 or WAV); Advanced Export (CSV, VTT, section timestamps)Text formats are free. Audio uses the MP3 export rate
More → Copy textCopies the current version as plain text (respects Show timestamps)Free
More → Show timestampsReader-only toggle for transcriptions. Hides · 0:34 headers without rewriting stored textFree
More → TranslateNew completed version in another language. Separate from New version → OriginalSame rate as MP3 export: 0.5 credits per started 30s of estimated speech (min 0.5). Refunded if translation fails
More → ShareUnlisted read-only link. Visitors see text only; no billed audioFree
More → Chat withOpens ChatGPT, Claude, Gemini, Perplexity, or Grok with a summarize prompt and the library URLFree (those products bill on their side)
More → Rename / folder / star / deleteRename, move, star, or permanently deleteFree

Create folders in Library before Move to folder has destinations.

Main column (read mode)

Speechdash splits the version into sentences, highlights the one being spoken, and lets you click a sentence to jump. Code blocks are skipped automatically; skip markers sit in the left gutter. Details: Sentence highlighting and Skip sentences.

Other main-column states

StateWhat you seePlayback bar
Edit modeLexical editor and floating selection toolbarHidden until you leave edit
PDF processingProgress panel while text is extractedHidden until the import completes
Import failedError panel with the failure messageHidden
Version generatingYou stay on the current version; the header spinner tracks the new oneUnchanged on the current version

PDF details: PDF import processing. Failed imports: PDF import failed.

Playback bar

Hidden while editing or while a PDF is still processing. On mobile it docks full-width at the bottom; on desktop it is a centered pill.

ControlBehaviorCredits
Previous / NextJump one spoken sentence (skipped sentences are omitted)Free
Play / PauseStarts or stops audio. Buffering shows a spinner on the play buttonEco: 0. Cloud: ~1 credit per minute at 1×; replays of sentences already cached on the server cost 0
TimelineScrub by sentence across the versionSame as play for Cloud sentences you have not heard yet
Speed0.5×–2.5× for this version, applied immediatelyCloud cost scales with spoken time, not wall-clock speed
Voice popoverPreset voice, listening language, Eco or CloudEco is local and free. Cloud uses the Cloud rate. Saved on this version
Voice load statusEco may show a model-download bar on first useFree (download, not wallet)
MicClick to record a question; click again to stop. Opens Ask questions and fills the box (not auto-sent). Pauses playbackSpeech-to-text for the question is not billed as Cloud playback
Ask questionsOpens or closes the right panelSending a question bills tokens + 1 credit per assistant reply

If Eco cannot run (mobile or missing WebAssembly), Cloud is used when available. If Cloud is down, the player can fall back to Eco. See Eco vs Cloud and Live playback customization.

Playback stops when you leave the reader.

Ask questions panel

Closed by default. Drag the divider to resize (about 18–50% of the width).

ControlBehaviorCredits
Message listGrounded in this document (retrieval). Enter inserts a newline; Send submitsTokens + 1 credit per assistant message
Speak-aloudAuto-plays each reply once; you can replay from the messageSame as document playback: Eco free, Cloud ~1 credit/min
Clear history (brush)Deletes saved messages for this fileFree; later questions send a smaller context
CloseHides the panel; history stays until you clear itFree

Guides: Ask questions, How asking questions works, Voice input, Assistant speak-aloud, Clear history.

Edit mode (selection toolbar)

Select text to format, skip, or enhance. Paste from chat or code editors is cleaned up automatically.

ActionCredits
Headings, bold, italic, strikethrough, code, clear formattingFree
Skip / Narrator on the selectionFree (changes what is spoken)
Enhance (optimize for audio, summarize, fix spelling)Tokens + 1 credit per run

Save re-segments sentences and can invalidate Cloud audio cache for changed lines. Unsaved edits: Unsaved changes.

Translate vs New version

PathResultBilling
More → TranslateA finished version in another language, named after that languageMP3-export rate on estimated speech; refund if translation fails
New version → Original with another output languageLanguage-model rewrite/copy into that languageTokens + 1 credit
New version Natural / Lecture / ConversationalAI listening stylesTokens + 1 credit

Share

More → Share publishes {origin}/share/document/{id}. Anyone with the link can read the current completed version. They cannot edit, ask questions, or play billed audio. Share a document.

Billing on this screen (summary)

ActionTypical cost
Eco listen, edit, copy, share, outline, timestamps, star, rename, folders, delete, text export0 credits
Cloud listen / Cloud speak-aloud~1 credit/min at 1×; cached sentence replay 0
Ask questionsTokens + 1 credit per reply
Enhance selectionTokens + 1 credit per run
New version (AI types / Original)Tokens + 1 credit
Translate0.5 credits per started 30s of estimated speech (min 0.5)
Download audio (MP3/WAV)0.5 credits per started 30s per sentence (min 0.5); Cloud-cached sentences 0

Wallet details: What uses credits. Out of credits: Out of credits.

Estimates in dialogs use the coins icon and match display credits in Settings → Billing.

On this page