Every Recitare feature explained: the complete guide

Drag and drop a file onto the upload area, or click to browse. Recitare accepts PDF, DOCX, PPTX, TXT, JPG, PNG, and TIFF files. PDFs are parsed directly in your browser, no upload to a server required for text extraction. Images are processed with Tesseract.js OCR, also in-browser.

You can also paste a URL (any webpage or arXiv paper link) and Recitare will extract the article text automatically. You can also search academic papers from arXiv and PubMed.

How it works: After parsing, non-PDF documents are sent through an AI text revalidation step that reorders text into the correct human reading sequence, fixing issues like cross-column merging in PowerPoint slides or incorrect text flow. You see the text immediately; the revalidated version silently replaces it within seconds.

When you press Play, each sentence highlights in real time as Recitare reads it aloud. The currently active sentence is emphasized with a coral highlight, while completed sentences fade to a lighter color. The reading area auto-scrolls to keep the active sentence centered.

This dual-channel approach, hearing and seeing the text simultaneously, is backed by multimedia learning theory (Mayer, 2009). Your brain processes auditory and visual information through separate channels, so combining them improves retention compared to reading or listening alone.

Tip: Choose “Flow State” above a PDF (or in Tools on a phone) for a clean, distraction-free reading experience with larger text and wider line spacing, ideal for long study sessions.

The speed selector is in the bottom playback bar, or Tools on a phone. Choose 0.75x, 1x, 1.25x, 1.5x, 1.75x or 2x. Changing speed never changes the pitch, so voices stay natural at every setting.

When to slow down: Dense methodology sections, statistical results, or unfamiliar terminology. Dropping to 0.75x gives your brain more processing time for complex material.

When to speed up: Literature reviews, background sections, or papers you're triaging. At 2x, a 30-minute paper becomes 15 minutes. Use it to decide which papers deserve a deep read.

Sprint Mode (Researcher): A 10-minute focused sprint with a timer and progress bar. The voice gets slightly more energetic and pacing increases. Great for overcoming procrastination. Commit to just 10 minutes.

Pick from 6 voice personas across 3 accents: American (male/female), British (male/female), and Australian (male/female). All six are available on every plan, powered by Inworld. Researcher uses Inworld's highest-quality model.

On the Researcher tier, voices gain expressive audio capabilities: they can sound enthusiastic, thoughtful, whispered, or analytical depending on the content. The voice adapts based on your chosen “vibe”: hyped for energy, focused for concentration, or analysis for deep reading.

Language support: Recitare speaks 14+ languages natively via Inworld. If your document is in French, German, Spanish, Japanese, or another supported language, Recitare detects it automatically and reads in that language.

Stuck on a passage? In Study mode, tap a passage and choose Explain or Plain English.

How it works: Your selected text is sent to Gemini with the full document as context, the whole of it. This means the explanation doesn't just paraphrase. It reflects how the passage fits into the paper's argument.

Free tier: 3 Quick Insights per day. Scholar and Researcher: As many as you need within fair use (fair use).

Open the chat panel and ask anything about your document: “What's the main argument in Section 3?”, “Summarize the methodology in two sentences”, or “Does Figure 2 support or contradict the hypothesis?”

Every answer includes page citations (e.g., [Page 4]) so you can verify claims against the source. Recitare uses context caching. After the first query, follow-up questions are faster and cheaper because the document is already in memory.

Ask While Listening (Scholar/Researcher): A floating “raise your hand” button appears during TTS playback. Tap it to pause audio and ask a question about what you just heard, then resume where you left off.

Free tier: 5 messages per day. Scholar and Researcher: As many as you need within fair use (fair use, soft cap at 1,000/month).

Click the Snap-to-Explain icon, then drag a selection rectangle over any figure, chart, table, or diagram in your PDF. Recitare captures it at 300 DPI (4x higher than screen resolution) for accurate AI analysis.

A scanning animation plays while Recitare analyzes the image. Within seconds, the Tutor Sidebar opens with the explanation: what the axes represent, the key trends, and what it means for the paper's argument. The explanation is narrated aloud automatically.

Persistent insights: Snapshot locations are saved. When you revisit the document, amber lightbulb icons appear over previously snapped areas. Click them to replay the explanation without re-analyzing.

Copy to Clipboard: Click “Copy to Clipboard” in the Tutor Sidebar to copy the screenshot + explanation to your clipboard as rich HTML. Paste directly into Goodnotes, Notion, or any note-taking app.

In Study mode, Recitare pauses after each section and asks a comprehension question about what you just read. These aren't generic trivia. They're Socratic probes designed to test whether you understood the argument, not just the facts.

Why it works: Active recall (retrieving information from memory) leads to better long-term retention than re-reading the same passage. By testing you during reading rather than after, Recitare catches misunderstandings before they compound.

Available on Researcher tier.

Generate a full document summary with one click. Recitare produces a structured summary that preserves the paper's logical flow and includes [Page X] citations for every claim.

Summaries can be exported as Markdown, copied to clipboard (rich HTML), or downloaded as PDF.

Free tier: 3 summaries per day. Scholar and Researcher: As many as you need, within fair use.

Choose from 5 study guide formats:

• Key Concepts: Core ideas distilled into bullet points
• Q&A Pairs: Question-and-answer format for self-testing
• Cornell Notes: Cues, notes, and summary in the classic layout
• Outline: Hierarchical structure of the document's argument
• Flashcards: Front/back cards for spaced repetition

All formats support export (Markdown download, clipboard copy, PDF).

Free tier: Key Concepts only. Scholar and Researcher: All 5 formats.

Converts your document into a two-voice podcast dialogue: two AI hosts discuss the paper's key findings, methods, and implications in a conversational format. A podcast made from your own paper, built into your reading workflow.

Includes a full transcript you can follow along with. Great for getting an overview before diving into the full text, or for reviewing material during a commute.

Available on Researcher tier (50 per month).

A client-side text analysis toolkit inspired by Voyant Tools. No API calls. Everything runs in your browser on the parsed text. Four tabs:

• Word Cloud: TF-IDF weighted terms, click to explore
• Term Heatmap: Per-page term frequency, click segments to jump to that page
• Concordance (KWIC): Every occurrence of a term with surrounding context
• Collocates: Words that statistically co-occur with your selected term

Designed for grad students doing literature reviews. Use it to triage papers and decide which ones deserve a deep read.

Available on all tiers (runs entirely in-browser).

For PDFs, choose between two reading layouts:

• Document: Original PDF pages with zoom and passage highlighting, preserving the document's visual structure.
• Flow State: Reflowed serif text with word-by-word highlighting, comfortable spacing and paper, sepia or dark themes. Other document formats use this reading layout.